Authors: Zhihuan Jiang, Zhen Yang, Jinhao Chen, Zhengxiao Du, Weihan Wang, Bin Xu, Yuxiao Dong, Jie Tang
Abstract: Multi-modal large language models (MLLMs) have demonstrated promising capabilities across various tasks by integrating textual and visual information to achieve visual understanding in complex scenarios. Despite the availability of several benchmarks aims to evaluating MLLMs in tasks from visual question answering to complex problem-solving, most focus predominantly on mathematics or general visual understanding tasks. This reveals a critical gap in current benchmarks, which often overlook the inclusion of other key scientific disciplines such as physics and chemistry. To address this gap, we meticulously construct a comprehensive benchmark, named VisScience, which is utilized to assess the multi-modal scientific reasoning across the three disciplines of mathematics, physics, and chemistry. This benchmark comprises 3,000 questions drawn from K12 education - spanning elementary school through high school - equally distributed across three disciplines, with 1,000 questions per discipline. The questions within VisScience span 21 distinct subjects and are categorized into five difficulty levels, offering a broad spectrum of topics within each discipline. With VisScience, we present a detailed evaluation of the performance of 25 representative MLLMs in scientific reasoning. Experimental results demonstrate that closed-source MLLMs generally outperform open-source models. The best performance observed include a 53.4\% accuracy in mathematics by Claude3.5-Sonnet, 38.2\% in physics by GPT-4o, and 47.0\% in chemistry by Gemini-1.5-Pro. These results underscore the strengths and limitations of MLLMs, suggesting areas for future improvement and highlighting the importance of developing models that can effectively handle the diverse demands of multi-modal scientific reasoning.
Authors: Gaurab Pokharel
Abstract: Prior studies have demonstrated that for many real-world problems, POMDPs can be solved through online algorithms both quickly and with near optimality. However, on an important set of problems where there is a large time delay between when the agent can gather information and when it needs to use that information, these solutions fail to adequately consider the value of information. As a result, information gathering actions, even when they are critical in the optimal policy, will be ignored by existing solutions, leading to sub-optimal decisions by the agent. In this research, we develop a novel solution that rectifies this problem by introducing a new algorithm that improves upon state-of-the-art online planning by better reflecting on the value of actions that gather information. We do this by adding Entropy to the UCB1 heuristic in the POMCP algorithm. We test this solution on the hallway problem. Results indicate that our new algorithm performs significantly better than POMCP.
Authors: Renato L. Conforme Rosado, Francisco C. Calderon Bocanegra
Abstract: This document presents the experimentation of 4 cases of food distribution for shrimp farming. The distributions are based on the location of the automatic feeders. Three cases applied in reality and a fourth case where the food is irrigated on the crop simultaneously and uniformly. In a first stage, the simulation of the three distribution cases is successfully adjusted to reality, where the trend of the shrimp growth curve is correlated with the historical data curve. A second stage where you experiment in 16 configurations that are based on the amount of food, the density of biomass and the distribution of the food. The simulation adopts the concepts of genetic algorithms to improve the population and fuzzy logic as an agent evaluation technique for decision-making against the quality of physical-chemical parameters in the simulated environment. The results of these interactions reveal a reduction in the simulated total culture time from 22 weeks to 14 weeks.
Authors: Mengyun Qiao, Kathryn A McGurk, Shuo Wang, Paul M. Matthews, Declan P O Regan, Wenjia Bai
Abstract: Understanding the structure and motion of the heart is crucial for diagnosing and managing cardiovascular diseases, the leading cause of global death. There is wide variation in cardiac shape and motion patterns, that are influenced by demographic, anthropometric and disease factors. Unravelling the normal patterns of shape and motion, as well as understanding how each individual deviates from the norm, would facilitate accurate diagnosis and personalised treatment strategies. To this end, we developed a novel conditional generative model, MeshHeart, to learn the distribution of cardiac shape and motion patterns. MeshHeart is capable of generating 3D+t cardiac mesh sequences, taking into account clinical factors such as age, sex, weight and height. To model the high-dimensional and complex spatio-temporal mesh data, MeshHeart employs a geometric encoder to represent cardiac meshes in a latent space, followed by a temporal Transformer to model the motion dynamics of latent representations. Based on MeshHeart, we investigate the latent space of 3D+t cardiac mesh sequences and propose a novel distance metric termed latent delta, which quantifies the deviation of a real heart from its personalised normative pattern in the latent space. In experiments using a large dataset of 38,309 subjects, MeshHeart demonstrates a high performance in cardiac mesh sequence reconstruction and generation. Features defined in the latent space are highly discriminative for cardiac disease classification, whereas the latent delta exhibits strong correlation with clinical phenotypes in phenome-wide association studies. The codes and models of this study will be released to benefit further research on digital heart modelling.
Authors: Ayoob Sadeghiani
Abstract: AI governance and ethics in AI development have become critical concerns, prompting active discussions among tech companies, governments, and researchers about the potential risks AI poses to our democracies. This short essay aims to highlight one such risk: how generative AI includes or excludes equity-deserving groups in its outputs. The findings reveal that generative AI is not equitably inclusive regarding gender, race, age, and visible disability.
Authors: Zhao Cheng, Diane Wan, Matthew Abueg, Sahra Ghalebikesabi, Ren Yi, Eugene Bagdasarian, Borja Balle, Stefan Mellem, Shawn O'Banion
Abstract: Advances in generative AI point towards a new era of personalized applications that perform diverse tasks on behalf of users. While general AI assistants have yet to fully emerge, their potential to share personal data raises significant privacy challenges. This paper introduces CI-Bench, a comprehensive synthetic benchmark for evaluating the ability of AI assistants to protect personal information during model inference. Leveraging the Contextual Integrity framework, our benchmark enables systematic assessment of information flow across important context dimensions, including roles, information types, and transmission principles. We present a novel, scalable, multi-step synthetic data pipeline for generating natural communications, including dialogues and emails. Unlike previous work with smaller, narrowly focused evaluations, we present a novel, scalable, multi-step data pipeline that synthetically generates natural communications, including dialogues and emails, which we use to generate 44 thousand test samples across eight domains. Additionally, we formulate and evaluate a naive AI assistant to demonstrate the need for further study and careful training towards personal assistant tasks. We envision CI-Bench as a valuable tool for guiding future language model development, deployment, system design, and dataset construction, ultimately contributing to the development of AI assistants that align with users' privacy expectations.
Authors: Jaewan Park, Shashank Kushwaha, Junyan He, Seid Koric, Qibang Liu, Iwona Jasiuk, Diab Abueidda
Abstract: Metamaterials, synthetic materials with customized properties, have emerged as a promising field due to advancements in additive manufacturing. These materials derive unique mechanical properties from their internal lattice structures, which are often composed of multiple materials that repeat geometric patterns. While traditional inverse design approaches have shown potential, they struggle to map nonlinear material behavior to multiple possible structural configurations. This paper presents a novel framework leveraging video diffusion models, a type of generative artificial Intelligence (AI), for inverse multi-material design based on nonlinear stress-strain responses. Our approach consists of two key components: (1) a fields generator using a video diffusion model to create solution fields based on target nonlinear stress-strain responses, and (2) a structure identifier employing two UNet models to determine the corresponding multi-material 2D design. By incorporating multiple materials, plasticity, and large deformation, our innovative design method allows for enhanced control over the highly nonlinear mechanical behavior of metamaterials commonly seen in real-world applications. It offers a promising solution for generating next-generation metamaterials with finely tuned mechanical characteristics.
Authors: Binxia Xu, Antonis Bikakis, Daniel Onah, Andreas Vlachidis, Luke Dickens
Abstract: Given that AI systems are set to play a pivotal role in future decision-making processes, their trustworthiness and reliability are of critical concern. Due to their scale and complexity, modern AI systems resist direct interpretation, and alternative ways are needed to establish trust in those systems, and determine how well they align with human values. We argue that good measures of the information processing similarities between AI and humans, may be able to achieve these same ends. While Representational alignment (RA) approaches measure similarity between the internal states of two systems, the associated data can be expensive and difficult to collect for human systems. In contrast, Behavioural alignment (BA) comparisons are cheaper and easier, but questions remain as to their sensitivity and reliability. We propose two new behavioural alignment metrics misclassification agreement which measures the similarity between the errors of two systems on the same instances, and class-level error similarity which measures the similarity between the error distributions of two systems. We show that our metrics correlate well with RA metrics, and provide complementary information to another BA metric, within a range of domains, and set the scene for a new approach to value alignment.
Authors: Nels Numan, Shwetha Rajaram, Balasaravanan Thoravi Kumaravel, Nicolai Marquardt, Andrew D. Wilson
Abstract: There is increased interest in using generative AI to create 3D spaces for Virtual Reality (VR) applications. However, today's models produce artificial environments, falling short of supporting collaborative tasks that benefit from incorporating the user's physical context. To generate environments that support VR telepresence, we introduce SpaceBlender, a novel pipeline that utilizes generative AI techniques to blend users' physical surroundings into unified virtual spaces. This pipeline transforms user-provided 2D images into context-rich 3D environments through an iterative process consisting of depth estimation, mesh alignment, and diffusion-based space completion guided by geometric priors and adaptive text prompts. In a preliminary within-subjects study, where 20 participants performed a collaborative VR affinity diagramming task in pairs, we compared SpaceBlender with a generic virtual environment and a state-of-the-art scene generation framework, evaluating its ability to create virtual spaces suitable for collaboration. Participants appreciated the enhanced familiarity and context provided by SpaceBlender but also noted complexities in the generative environments that could detract from task focus. Drawing on participant feedback, we propose directions for improving the pipeline and discuss the value and design of blended spaces for different scenarios.
Authors: Bridget Leonard, Kristin Woodard, Scott O. Murray
Abstract: This study extends previous research on spatial representations in multimodal AI systems. Although current models demonstrate a rich understanding of spatial information from images, this information is rooted in propositional representations, which differ from the analog representations employed in human and animal spatial cognition. To further explore these limitations, we apply techniques from cognitive and developmental science to assess the perspective-taking abilities of GPT-4o. Our analysis enables a comparison between the cognitive development of the human brain and that of multimodal AI, offering guidance for future research and model development.
Authors: Aditya Singh, Haohan Wang
Abstract: As per recent studies, Self-supervised learning (SSL) does not readily extend to smaller architectures. One direction to mitigate this shortcoming while simultaneously training a smaller network without labels is to adopt unsupervised knowledge distillation (UKD). Existing UKD approaches handcraft preservation worthy inter/intra sample relationships between the teacher and its student. However, this may overlook/ignore other key relationships present in the mapping of a teacher. In this paper, instead of heuristically constructing preservation worthy relationships between samples, we directly motivate the student to model the teacher's embedding manifold. If the mapped manifold is similar, all inter/intra sample relationships are indirectly conserved. We first demonstrate that prior methods cannot preserve teacher's latent manifold due to their sole reliance on $L_2$ normalised embedding features. Subsequently, we propose a simple objective to capture the lost information due to normalisation. Our proposed loss component, termed \textbf{space similarity}, motivates each dimension of a student's feature space to be similar to the corresponding dimension of its teacher. We perform extensive experiments demonstrating strong performance of our proposed approach on various benchmarks.
Authors: Vu Tuan Truong, Long Bao Le
Abstract: Diffusion models (DMs) are advanced deep learning models that achieved state-of-the-art capability on a wide range of generative tasks. However, recent studies have shown their vulnerability regarding backdoor attacks, in which backdoored DMs consistently generate a designated result (e.g., a harmful image) called backdoor target when the models' input contains a backdoor trigger. Although various backdoor techniques have been investigated to attack DMs, defense methods against these threats are still limited and underexplored, especially in inverting the backdoor trigger. In this paper, we introduce PureDiffusion, a novel backdoor defense framework that can efficiently detect backdoor attacks by inverting backdoor triggers embedded in DMs. Our extensive experiments on various trigger-target pairs show that PureDiffusion outperforms existing defense methods with a large gap in terms of fidelity (i.e., how much the inverted trigger resembles the original trigger) and backdoor success rate (i.e., the rate that the inverted trigger leads to the corresponding backdoor target). Notably, in certain cases, backdoor triggers inverted by PureDiffusion even achieve higher attack success rate than the original triggers.
Authors: Jin Du, Xinhe Zhang, Hao Shen, Xun Xian, Ganghua Wang, Jiawei Zhang, Yuhong Yang, Na Li, Jia Liu, Jie Ding
Abstract: Lifelong learning in artificial intelligence (AI) aims to mimic the biological brain's ability to continuously learn and retain knowledge, yet it faces challenges such as catastrophic forgetting. Recent neuroscience research suggests that neural activity in biological systems undergoes representational drift, where neural responses evolve over time, even with consistent inputs and tasks. We hypothesize that representational drift can alleviate catastrophic forgetting in AI during new task acquisition. To test this, we introduce DriftNet, a network designed to constantly explore various local minima in the loss landscape while dynamically retrieving relevant tasks. This approach ensures efficient integration of new information and preserves existing knowledge. Experimental studies in image classification and natural language processing demonstrate that DriftNet outperforms existing models in lifelong learning. Importantly, DriftNet is scalable in handling a sequence of tasks such as sentiment analysis and question answering using large language models (LLMs) with billions of parameters on a single Nvidia A100 GPU. DriftNet efficiently updates LLMs using only new data, avoiding the need for full dataset retraining. Tested on GPT-2 and RoBERTa, DriftNet is a robust, cost-effective solution for lifelong learning in LLMs. This study not only advances AI systems to emulate biological learning, but also provides insights into the adaptive mechanisms of biological neural systems, deepening our understanding of lifelong learning in nature.
Authors: Zhangcheng Qiang, Kerry Taylor, Weiqing Wang, Jing Jiang
Abstract: Hallucinations of large language models (LLMs) commonly occur in domain-specific downstream tasks, with no exception in ontology matching (OM). The prevalence of using LLMs for OM raises the need for benchmarks to better understand LLM hallucinations. The OAEI-LLM dataset is an extended version of the Ontology Alignment Evaluation Initiative (OAEI) datasets that evaluate LLM-specific hallucinations in OM tasks. We outline the methodology used in dataset construction and schema extension, and provide examples of potential use cases.
Authors: Goran Buba\v{s}
Abstract: OpenAI's ChatGPT (GPT-4 and GPT-4o) and other Large Language Models (LLMs) like Microsoft's Copilot, Google's Gemini 1.5 Pro, and Antrophic's Claude 3.5 Sonnet can be effectively used in various phases of scientific research. Their performance in diverse verbal tasks and reasoning is close to or above the average human level and rapidly increasing, providing those models with a capacity that resembles a relatively high level of theory of mind. The current ability of LLMs to process information about human psychology and communication creates an opportunity for their scientific use in the fields of personality psychology and interpersonal communication skills. This article illustrates the possible uses of GPT-4o and other advanced LLMs for typical tasks in designing self-assessment scales for interpersonal communication skills measurement like the selection and improvement of scale items and evaluation of content validity of scales. The potential for automated item generation and application is illustrated as well. The case study examples are accompanied by prompts for LLMs that can be useful for these purposes. Finally, a summary is provided of the potential benefits of using LLMs in the process of evaluation, design, and improvement of interpersonal communication skills self-assessment scales.
Authors: Amrit Diggavi Seshadri
Abstract: With the size and cost of large transformer-based language models growing, recently, there has been interest in shortcut casting of early transformer hidden-representations to final-representations for cheaper model inference. In particular, shortcutting pre-trained transformers with linear transformations over early layers has been shown to improve precision in early inference. However, for large language models, even this becomes computationally expensive. In this work, we propose Narrow Jump to Conclusions (NJTC) and Normalized Narrow Jump to Conclusions (N-NJTC) - parameter efficient alternatives to standard linear shortcutting that reduces shortcut parameter count by over 97%. We show that N-NJTC reliably outperforms Identity shortcuts at early stages and offers stable precision from all transformer block levels for GPT-2-XL, Phi3-Mini and Llama2-7B transformer models, demonstrating the viability of more parameter efficient short-cutting approaches.
Authors: Yibo Li, Yuan Fang, Mengmei Zhang, Chuan Shi
Abstract: Understanding molecular structure and related knowledge is crucial for scientific research. Recent studies integrate molecular graphs with their textual descriptions to enhance molecular representation learning. However, they focus on the whole molecular graph and neglect frequently occurring subgraphs, known as motifs,which are essential for determining molecular properties. Without such fine-grained knowledge, these models struggle to generalize to unseen molecules and tasks that require motif-level insights. To bridge this gap, we propose FineMolTex, a novel Fine-grained Molecular graph-Text pre-training framework to jointly learn coarse-grained molecule-level knowledge and fine-grained motif-level knowledge. Specifically, FineMolTex consists of two pre-training tasks: a contrastive alignment task for coarse-grained matching and a masked multi-modal modeling task for fine-grained matching. In particular, the latter predicts the labels of masked motifs and words, leveraging insights from each other, thereby enabling FineMolTex to understand the fine-grained matching between motifs and words. Finally, we conduct extensive experiments across three downstream tasks, achieving up to 230% improvement in the text-based molecule editing task. Additionally, our case studies reveal that FineMolTex successfully captures fine-grained knowledge, potentially offering valuable insights for drug discovery and catalyst design.
Authors: Yuxuan Zhua, Shiyi Wang, Wenqing Zhong, Nianchen Shen, Yunqi Li, Siqi Wang, Zhiheng Li, Cathy Wu, Zhengbing He, Li Li
Abstract: Artificial intelligence (AI) plays a crucial role in autonomous driving (AD) research, propelling its development towards intelligence and efficiency. Currently, the development of AD technology follows two main technical paths: modularization and end-to-end. Modularization decompose the driving task into modules such as perception, prediction, planning, and control, and train them separately. Due to the inconsistency of training objectives between modules, the integrated effect suffers from bias. End-to-end attempts to address this issue by utilizing a single model that directly maps from sensor data to control signals. This path has limited learning capabilities in a comprehensive set of features and struggles to handle unpredictable long-tail events and complex urban traffic scenarios. In the face of challenges encountered in both paths, many researchers believe that large language models (LLMs) with powerful reasoning capabilities and extensive knowledge understanding may be the solution, expecting LLMs to provide AD systems with deeper levels of understanding and decision-making capabilities. In light of the challenges faced by both paths, many researchers believe that LLMs, with their powerful reasoning abilities and extensive knowledge, could offer a solution. To understand if LLMs could enhance AD, this paper conducts a thorough analysis of the potential applications of LLMs in AD systems, including exploring their optimization strategies in both modular and end-to-end approaches, with a particular focus on how LLMs can tackle the problems and challenges present in current solutions. Furthermore, we discuss an important question: Can LLM-based artificial general intelligence (AGI) be a key to achieve high-level AD? We further analyze the potential limitations and challenges that LLMs may encounter in promoting the development of AD technology.
Authors: Carlos Pambo, Jacomine Grobler
Abstract: The use of trucks and drones as a solution to address last-mile delivery challenges is a new and promising research direction explored in this paper. The variation of the problem where the drone can intercept the truck while in movement or at the customer location is part of an optimisation problem called the vehicle routing problem with drones with interception (VRPDi). This paper proposes an evolutionary algorithm to solve the VRPDi. In this variation of the VRPDi, multiple pairs of trucks and drones need to be scheduled. The pairs leave and return to a depot location together or separately to make deliveries to customer nodes. The drone can intercept the truck after the delivery or meet up with the truck at the following customer location. The algorithm was executed on the travelling salesman problem with drones (TSPD) datasets by Bouman et al. (2015), and the performance of the algorithm was compared by benchmarking the results of the VRPDi against the results of the VRP of the same dataset. This comparison showed improvements in total delivery time between 39% and 60%. Further detailed analysis of the algorithm results examined the total delivery time, distance, node delivery scheduling and the degree of diversity during the algorithm execution. This analysis also considered how the algorithm handled the VRPDi constraints. The results of the algorithm were then benchmarked against algorithms in Dillon et al. (2023) and Ernst (2024). The latter solved the problem with a maximum drone distance constraint added to the VRPDi. The analysis and benchmarking of the algorithm results showed that the algorithm satisfactorily solved 50 and 100-nodes problems in a reasonable amount of time, and the solutions found were better than those found by the algorithms in Dillon et al. (2023) and Ernst (2024) for the same problems.
Authors: Andre de Carvalho, Robson Bonidia, Jude Dzevela Kong, Mariana Dauhajre, Claudio Struchiner, Guilherme Goedert, Peter F. Stadler, Maria Emilia Walter, Danilo Sanches, Troy Day, Marcia Castro, John Edmunds, Manuel Colome-Hidalgo, Demian Arturo Herrera Morban, Edian F. Franco, Cesar Ugarte-Gil, Patricia Espinoza-Lopez, Gabriel Carrasco-Escobar, Ulisses Rocha
Abstract: Infectious diseases, transmitted directly or indirectly, are among the leading causes of epidemics and pandemics. Consequently, several open challenges exist in predicting epidemic outbreaks, detecting variants, tracing contacts, discovering new drugs, and fighting misinformation. Artificial Intelligence (AI) can provide tools to deal with these scenarios, demonstrating promising results in the fight against the COVID-19 pandemic. AI is becoming increasingly integrated into various aspects of society. However, ensuring that AI benefits are distributed equitably and that they are used responsibly is crucial. Multiple countries are creating regulations to address these concerns, but the borderless nature of AI requires global cooperation to define regulatory and guideline consensus. Considering this, The Global South AI for Pandemic & Epidemic Preparedness & Response Network (AI4PEP) has developed an initiative comprising 16 projects across 16 countries in the Global South, seeking to strengthen equitable and responsive public health systems that leverage Southern-led responsible AI solutions to improve prevention, preparedness, and response to emerging and re-emerging infectious disease outbreaks. This opinion introduces our branches in Latin American and Caribbean (LAC) countries and discusses AI governance in LAC in the light of biotechnology. Our network in LAC has high potential to help fight infectious diseases, particularly in low- and middle-income countries, generating opportunities for the widespread use of AI techniques to improve the health and well-being of their communities.
Authors: Sejin Kim, Hosung Lee, Sundong Kim
Abstract: The effectiveness of AI model training hinges on the quality of the trajectory data used, particularly in aligning the model's decision with human intentions. However, in the human task-solving trajectories, we observe significant misalignments between human intentions and the recorded trajectories, which can undermine AI model training. This paper addresses the challenges of these misalignments by proposing a visualization tool and a heuristic algorithm designed to detect and categorize discrepancies in trajectory data. Although the heuristic algorithm requires a set of predefined human intentions to function, which we currently cannot extract, the visualization tool offers valuable insights into the nature of these misalignments. We expect that eliminating these misalignments could significantly improve the utility of trajectory data for AI model training. We also propose that future work should focus on developing methods, such as Topic Modeling, to accurately extract human intentions from trajectory data, thereby enhancing the alignment between user actions and AI learning processes.
Authors: Kei-Sing Ng, Qingchen Wang
Abstract: The success of large-scale language models like GPT can be attributed to their ability to efficiently predict the next token in a sequence. However, these models rely on constant computational effort regardless of the complexity of the token they are predicting, lacking the capacity for iterative refinement. In this paper, we introduce a novel Loop-Residual Neural Network, which achieves better performance by utilizing longer computational time without increasing the model size. Our approach revisits the input multiple times, refining the prediction by iteratively looping over a subset of the model with residual connections. We demonstrate the effectiveness of this method through experiments comparing versions of GPT-2 with our Loop-Residual models, showing improved performance in language modeling tasks while maintaining similar parameter counts. Importantly, these improvements are achieved without the need for extra training data.
Authors: Oliver Bensch, Leonie Bensch, Tommy Nilsson, Florian Saling, Bernd Bewer, Sophie Jentzsch, Tobias Hecking, J. Nathan Kutz
Abstract: This paper describes the capabilities and potential of the intelligent personal assistant (IPA) CORE (Checklist Organizer for Research and Exploration), designed to support astronauts during procedures onboard the International Space Station (ISS), the Lunar Gateway station, and beyond. We reflect on the importance of a reliable and flexible assistant capable of offline operation and highlight the usefulness of audiovisual interaction using augmented reality elements to intuitively display checklist information. We argue that current approaches to the design of IPAs in space operations fall short of meeting these criteria. Therefore, we propose CORE as an assistant that combines Knowledge Graphs (KGs), Retrieval-Augmented Generation (RAG) for a Generative Pre-Trained Transformer (GPT), and Augmented Reality (AR) elements to ensure an intuitive understanding of procedure steps, reliability, offline availability, and flexibility in terms of response style and procedure updates.
Authors: Jamal Al-Karaki, Philip Ilono, Sanchit Baweja, Jalal Naghiyev, Raja Singh Yadav, Muhammad Al-Zafar Khan
Abstract: Coronary Heart Disease affects millions of people worldwide and is a well-studied area of healthcare. There are many viable and accurate methods for the diagnosis and prediction of heart disease, but they have limiting points such as invasiveness, late detection, or cost. Supervised learning via machine learning algorithms presents a low-cost (computationally speaking), non-invasive solution that can be a precursor for early diagnosis. In this study, we applied several well-known methods and benchmarked their performance against each other. It was found that Random Forest with oversampling of the predictor variable produced the highest accuracy of 84%.
Authors: Yew Ken Chia, Qi Sun, Lidong Bing, Soujanya Poria
Abstract: Large multimodal models have demonstrated impressive problem-solving abilities in vision and language tasks, and have the potential to encode extensive world knowledge. However, it remains an open challenge for these models to perceive, reason, plan, and act in realistic environments. In this work, we introduce Can-Do, a benchmark dataset designed to evaluate embodied planning abilities through more diverse and complex scenarios than previous datasets. Our dataset includes 400 multimodal samples, each consisting of natural language user instructions, visual images depicting the environment, state changes, and corresponding action plans. The data encompasses diverse aspects of commonsense knowledge, physical understanding, and safety awareness. Our fine-grained analysis reveals that state-of-the-art models, including GPT-4V, face bottlenecks in visual perception, comprehension, and reasoning abilities. To address these challenges, we propose NeuroGround, a neurosymbolic framework that first grounds the plan generation in the perceived environment states and then leverages symbolic planning engines to augment the model-generated plans. Experimental results demonstrate the effectiveness of our framework compared to strong baselines. Our code and dataset are available at https://embodied-planning.github.io.
Authors: Naoki Yokoyama, Ram Ramrakhya, Abhishek Das, Dhruv Batra, Sehoon Ha
Abstract: We present the Habitat-Matterport 3D Open Vocabulary Object Goal Navigation dataset (HM3D-OVON), a large-scale benchmark that broadens the scope and semantic range of prior Object Goal Navigation (ObjectNav) benchmarks. Leveraging the HM3DSem dataset, HM3D-OVON incorporates over 15k annotated instances of household objects across 379 distinct categories, derived from photo-realistic 3D scans of real-world environments. In contrast to earlier ObjectNav datasets, which limit goal objects to a predefined set of 6-20 categories, HM3D-OVON facilitates the training and evaluation of models with an open-set of goals defined through free-form language at test-time. Through this open-vocabulary formulation, HM3D-OVON encourages progress towards learning visuo-semantic navigation behaviors that are capable of searching for any object specified by text in an open-vocabulary manner. Additionally, we systematically evaluate and compare several different types of approaches on HM3D-OVON. We find that HM3D-OVON can be used to train an open-vocabulary ObjectNav agent that achieves both higher performance and is more robust to localization and actuation noise than the state-of-the-art ObjectNav approach. We hope that our benchmark and baseline results will drive interest in developing embodied agents that can navigate real-world spaces to find household objects specified through free-form language, taking a step towards more flexible and human-like semantic visual navigation. Code and videos available at: naoki.io/ovon.
Authors: Ting Yu Tsai, Li Lin, Shu Hu, Connie W. Tsao, Xin Li, Ming-Ching Chang, Hongtu Zhu, Xin Wang
Abstract: Building on the success of deep learning models in cardiovascular structure segmentation, increasing attention has been focused on improving generalization and robustness, particularly in small, annotated datasets. Despite recent advancements, current approaches often face challenges such as overfitting and accuracy limitations, largely due to their reliance on large datasets and narrow optimization techniques. This paper introduces the UU-Mamba model, an extension of the U-Mamba architecture, designed to address these challenges in both cardiac and vascular segmentation. By incorporating Sharpness-Aware Minimization (SAM), the model enhances generalization by targeting flatter minima in the loss landscape. Additionally, we propose an uncertainty-aware loss function that combines region-based, distribution-based, and pixel-based components to improve segmentation accuracy by capturing both local and global features. While the UU-Mamba model has already demonstrated great performance, further testing is required to fully assess its generalization and robustness. We expand our evaluation by conducting new trials on the ImageCAS (coronary artery) and Aorta (aortic branches and zones) datasets, which present more complex segmentation challenges than the ACDC dataset (left and right ventricles) used in our previous work, showcasing the model's adaptability and resilience. We confirm UU-Mamba's superior performance over leading models such as TransUNet, Swin-Unet, nnUNet, and nnFormer. Moreover, we provide a more comprehensive evaluation of the model's robustness and segmentation accuracy, as demonstrated by extensive experiments.
Authors: Fariza Rashid, Nishavi Ranaweera, Ben Doyle, Suranga Seneviratne
Abstract: Malicious URL classification represents a crucial aspect of cyber security. Although existing work comprises numerous machine learning and deep learning-based URL classification models, most suffer from generalisation and domain-adaptation issues arising from the lack of representative training datasets. Furthermore, these models fail to provide explanations for a given URL classification in natural human language. In this work, we investigate and demonstrate the use of Large Language Models (LLMs) to address this issue. Specifically, we propose an LLM-based one-shot learning framework that uses Chain-of-Thought (CoT) reasoning to predict whether a given URL is benign or phishing. We evaluate our framework using three URL datasets and five state-of-the-art LLMs and show that one-shot LLM prompting indeed provides performances close to supervised models, with GPT 4-Turbo being the best model, followed by Claude 3 Opus. We conduct a quantitative analysis of the LLM explanations and show that most of the explanations provided by LLMs align with the post-hoc explanations of the supervised classifiers, and the explanations have high readability, coherency, and informativeness.
Authors: Ansh Chaurasia
Abstract: The presiding model generation algorithms rely on simple, inflexible adapter selection to provide personalized results. We propose the model-adapter composition problem as a generalized problem to past work factoring in practical hardware and affordability constraints, and introduce MANTA as a new approach to the problem. Experiments on COCO 2014 validation show MANTA to be superior in image task diversity and quality at the cost of a modest drop in alignment. Our system achieves a $94\%$ win rate in task diversity and a $80\%$ task quality win rate versus the best known system, and demonstrates strong potential for direct use in synthetic data generation and the creative art domains.
Authors: Gaole He, Abri Bharos, Ujwal Gadiraju
Abstract: Powerful predictive AI systems have demonstrated great potential in augmenting human decision making. Recent empirical work has argued that the vision for optimal human-AI collaboration requires 'appropriate reliance' of humans on AI systems. However, accurately estimating the trustworthiness of AI advice at the instance level is quite challenging, especially in the absence of performance feedback pertaining to the AI system. In practice, the performance disparity of machine learning models on out-of-distribution data makes the dataset-specific performance feedback unreliable in human-AI collaboration. Inspired by existing literature on critical thinking and a critical mindset, we propose the use of debugging an AI system as an intervention to foster appropriate reliance. In this paper, we explore whether a critical evaluation of AI performance within a debugging setting can better calibrate users' assessment of an AI system and lead to more appropriate reliance. Through a quantitative empirical study (N = 234), we found that our proposed debugging intervention does not work as expected in facilitating appropriate reliance. Instead, we observe a decrease in reliance on the AI system after the intervention -- potentially resulting from an early exposure to the AI system's weakness. We explore the dynamics of user confidence and user estimation of AI trustworthiness across groups with different performance levels to help explain how inappropriate reliance patterns occur. Our findings have important implications for designing effective interventions to facilitate appropriate reliance and better human-AI collaboration.
Authors: Chen Tessler, Yunrong Guo, Ofir Nabati, Gal Chechik, Xue Bin Peng
Abstract: Crafting a single, versatile physics-based controller that can breathe life into interactive characters across a wide spectrum of scenarios represents an exciting frontier in character animation. An ideal controller should support diverse control modalities, such as sparse target keyframes, text instructions, and scene information. While previous works have proposed physically simulated, scene-aware control models, these systems have predominantly focused on developing controllers that each specializes in a narrow set of tasks and control modalities. This work presents MaskedMimic, a novel approach that formulates physics-based character control as a general motion inpainting problem. Our key insight is to train a single unified model to synthesize motions from partial (masked) motion descriptions, such as masked keyframes, objects, text descriptions, or any combination thereof. This is achieved by leveraging motion tracking data and designing a scalable training method that can effectively utilize diverse motion descriptions to produce coherent animations. Through this process, our approach learns a physics-based controller that provides an intuitive control interface without requiring tedious reward engineering for all behaviors of interest. The resulting controller supports a wide range of control modalities and enables seamless transitions between disparate tasks. By unifying character control through motion inpainting, MaskedMimic creates versatile virtual characters. These characters can dynamically adapt to complex scenes and compose diverse motions on demand, enabling more interactive and immersive experiences.
Authors: Le Yang, Ziwei Zheng, Yizeng Han, Shiji Song, Gao Huang, Fan Li
Abstract: Differentiable architecture search (DARTS) has emerged as a promising technique for effective neural architecture search, and it mainly contains two steps to find the high-performance architecture: First, the DARTS supernet that consists of mixed operations will be optimized via gradient descent. Second, the final architecture will be built by the selected operations that contribute the most to the supernet. Although DARTS improves the efficiency of NAS, it suffers from the well-known degeneration issue which can lead to deteriorating architectures. Existing works mainly attribute the degeneration issue to the failure of its supernet optimization, while little attention has been paid to the selection method. In this paper, we cease to apply the widely-used magnitude-based selection method and propose a novel criterion based on operation strength that estimates the importance of an operation by its effect on the final loss. We show that the degeneration issue can be effectively addressed by using the proposed criterion without any modification of supernet optimization, indicating that the magnitude-based selection method can be a critical reason for the instability of DARTS. The experiments on NAS-Bench-201 and DARTS search spaces show the effectiveness of our method.
Authors: Daan Roordink, Sibylle Hess
Abstract: Probabilistic regression models trained with maximum likelihood estimation (MLE), can sometimes overestimate variance to an unacceptable degree. This is mostly problematic in the multivariate domain. While univariate models often optimize the popular Continuous Ranked Probability Score (CRPS), in the multivariate domain, no such alternative to MLE has yet been widely accepted. The Energy Score - the most investigated alternative - notoriously lacks closed-form expressions and sensitivity to the correlation between target variables. In this paper, we propose Conditional CRPS: a multivariate strictly proper scoring rule that extends CRPS. We show that closed-form expressions exist for popular distributions and illustrate their sensitivity to correlation. We then show in a variety of experiments on both synthetic and real data, that Conditional CRPS often outperforms MLE, and produces results comparable to state-of-the-art non-parametric models, such as Distributional Random Forest (DRF).
Authors: Yuntao Wang, Yanghe Pan, Quan Zhao, Yi Deng, Zhou Su, Linkang Du, Tom H. Luan
Abstract: Large Model (LM) agents, powered by large foundation models such as GPT-4 and DALL-E 2, represent a significant step towards achieving Artificial General Intelligence (AGI). LM agents exhibit key characteristics of autonomy, embodiment, and connectivity, allowing them to operate across physical, virtual, and mixed-reality environments while interacting seamlessly with humans, other agents, and their surroundings. This paper provides a comprehensive survey of the state-of-the-art in LM agents, focusing on the architecture, cooperation paradigms, security, privacy, and future prospects. Specifically, we first explore the foundational principles of LM agents, including general architecture, key components, enabling technologies, and modern applications. Then, we discuss practical collaboration paradigms from data, computation, and knowledge perspectives towards connected intelligence of LM agents. Furthermore, we systematically analyze the security vulnerabilities and privacy breaches associated with LM agents, particularly in multi-agent settings. We also explore their underlying mechanisms and review existing and potential countermeasures. Finally, we outline future research directions for building robust and secure LM agent ecosystems.
Authors: Yuri Gurevich, Andreas Blass
Abstract: A hundred years ago, logic was almost synonymous with foundational studies. The ongoing AI revolution raises many deep foundational problems involving neuroscience, philosophy, computer science, and logic. The goal of the following dialog is to provoke young logicians with a taste for foundations to notice the foundational problems raised by the AI revolution.
Authors: Yuxing Zhi, Yuan Guo, Kai Yuan, Hesong Wang, Heng Xu, Haina Yao, Albert C Yang, Guangrui Huang, Yuping Duan
Abstract: Background: Large language models (LLMs) have seen extraordinary advances with applications in clinical decision support. However, high-quality evidence is urgently needed on the potential and limitation of LLMs in providing accurate clinical decisions based on real-world medical data. Objective: To evaluate quantitatively whether universal state-of-the-art LLMs (ChatGPT and GPT-4) can predict the incidence risk of myocardial infarction (MI) with logical inference, and to further make comparison between various models to assess the performance of LLMs comprehensively. Methods: In this retrospective cohort study, 482,310 participants recruited from 2006 to 2010 were initially included in UK Biobank database and later on resampled into a final cohort of 690 participants. For each participant, tabular data of the risk factors of MI were transformed into standardized textual descriptions for ChatGPT recognition. Responses were generated by asking ChatGPT to select a score ranging from 0 to 10 representing the risk. Chain of Thought (CoT) questioning was used to evaluate whether LLMs make prediction logically. The predictive performance of ChatGPT was compared with published medical indices, traditional machine learning models and other large language models. Conclusions: Current LLMs are not ready to be applied in clinical medicine fields. Future medical LLMs are suggested to be expert in medical domain knowledge to understand both natural languages and quantified medical data, and further make logical inferences.
Authors: Yuri Gurevich
Abstract: The Fall 2024 Logic in Computer Science column of the Bulletin of EATCS is a little discussion on intelligence, measuring intelligence, and related issues, provoked by a fascinating must-read article ``On the measure of intelligence'' by Fran\c{c}ois Chollet. The discussion includes a modicum of critique of the article.
Authors: Shaowei Ying, Zhenlong Li, Manzhu Yu
Abstract: The resurgence and rapid advancement of Generative Artificial Intelligence (GenAI) in 2023 has catalyzed transformative shifts across numerous industry sectors, including urban transportation and logistics. This study investigates the evaluation of Large Language Models (LLMs), specifically GPT-4 and Phi-3-mini, to enhance transportation planning. The study assesses the performance and spatial comprehension of these models through a transportation-informed evaluation framework that includes general geospatial skills, general transportation domain skills, and real-world transportation problem-solving. Utilizing a mixed-methods approach, the research encompasses an evaluation of the LLMs' general Geographic Information System (GIS) skills, general transportation domain knowledge as well as abilities to support human decision-making in the real-world transportation planning scenarios of congestion pricing. Results indicate that GPT-4 demonstrates superior accuracy and reliability across various GIS and transportation-specific tasks compared to Phi-3-mini, highlighting its potential as a robust tool for transportation planners. Nonetheless, Phi-3-mini exhibits competence in specific analytical scenarios, suggesting its utility in resource-constrained environments. The findings underscore the transformative potential of GenAI technologies in urban transportation planning. Future work could explore the application of newer LLMs and the impact of Retrieval-Augmented Generation (RAG) techniques, on a broader set of real-world transportation planning and operations challenges, to deepen the integration of advanced AI models in transportation management practices.
Authors: Michael Timothy Bennett, Sean Welsh, Anna Ciaunica
Abstract: We tackle the hard problem of consciousness taking the naturally-selected, self-organising, embodied organism as our starting point. We provide a mathematical formalism describing how biological systems self-organise to hierarchically interpret unlabelled sensory information according to valence and specific needs. Such interpretations imply behavioural policies which can only be differentiated from each other by the qualitative aspect of information processing. Selection pressures favour systems that can intervene in the world to achieve homeostatic and reproductive goals. Quality is a property arising in such systems to link cause to affect to motivate real world interventions. This produces a range of qualitative classifiers (interoceptive and exteroceptive) that motivate specific actions and determine priorities and preferences. Building upon the seminal distinction between access and phenomenal consciousness, our radical claim here is that phenomenal consciousness without access consciousness is likely very common, but the reverse is implausible. To put it provocatively: Nature does not like zombies. We formally describe the multilayered architecture of self-organisation from rocks to Einstein, illustrating how our argument applies in the real world. We claim that access consciousness at the human level is impossible without the ability to hierarchically model i) the self, ii) the world/others and iii) the self as modelled by others. Phenomenal consciousness is therefore required for human-level functionality. Our proposal lays the foundations of a formal science of consciousness, deeply connected with natural selection rather than abstract thinking, closer to human fact than zombie fiction.
Authors: Ildar Rakhmatulin
Abstract: This paper introduces a novel method for effectively removing artifacts from EEG signals by combining the Empirical Mode Decomposition (EMD) method with a machine learning architecture. The proposed method addresses the limitations of existing artifact removal techniques by enhancing the EMD method through interpolation of the upper and lower. For conventional artifact removal methods, the EMD technique is commonly employed. However, the challenge lies in accurately interpolating the missing components of the signal while preserving its inherent frequency components. To overcome this limitation, we incorporated machine learning technique, which enables us to carefully handle the interpolation process without directly manipulating the data. The key advantage of our approach lies in the preservation of the natural characteristics of the EEG signal during artifact removal. By utilizing machine learning for interpolation, we ensure that the average component obtained through the EMD method retains the crucial frequency components of the original signal. This preservation is essential for maintaining the integrity and fidelity of the EEG data, allowing for accurate analysis and interpretation. The results obtained from our evaluation serve to validate the effectiveness of our approach and pave the way for further advancements in EEG signal processing and analysis.
Authors: Vishal Mirza, Rahul Kulkarni, Aakanksha Jadhav
Abstract: Recent advancements in Large Language Models(LLMs) have been notable, yet widespread enterprise adoption remains limited due to various constraints. This paper examines bias in LLMs-a crucial issue affecting their usability, reliability, and fairness. Researchers are developing strategies to mitigate bias, including debiasing layers, specialized reference datasets like Winogender and Winobias, and reinforcement learning with human feedback (RLHF). These techniques have been integrated into the latest LLMs. Our study evaluates gender bias in occupational scenarios and gender, age, and racial bias in crime scenarios across four leading LLMs released in 2024: Gemini 1.5 Pro, Llama 3 70B, Claude 3 Opus, and GPT-4o. Findings reveal that LLMs often depict female characters more frequently than male ones in various occupations, showing a 37% deviation from US BLS data. In crime scenarios, deviations from US FBI data are 54% for gender, 28% for race, and 17% for age. We observe that efforts to reduce gender and racial bias often lead to outcomes that may over-index one sub-class, potentially exacerbating the issue. These results highlight the limitations of current bias mitigation techniques and underscore the need for more effective approaches.
Authors: Michele Laurelli
Abstract: As large-scale AI systems proliferate, ensuring compliance with data privacy laws such as the General Data Protection Regulation (GDPR) has become critical. This paper introduces Brain Surgery, a transformative methodology for making every local AI model GDPR-ready by enabling real-time privacy management and targeted unlearning. Building on advanced techniques such as Embedding-Corrupted Prompts (ECO Prompts), blockchain-based privacy management, and privacy-aware continual learning, Brain Surgery provides a modular solution that can be deployed across various AI architectures. This tool not only ensures compliance with privacy regulations but also empowers users to define their own privacy limits, creating a new paradigm in AI ethics and governance.
Authors: Huayun Zhang, Jeremy H. M. Wong, Geyu Lin, Nancy F. Chen
Abstract: Speech evaluation measures a learners oral proficiency using automatic models. Corpora for training such models often pose sparsity challenges given that there often is limited scored data from teachers, in addition to the score distribution across proficiency levels being often imbalanced among student cohorts. Automatic scoring is thus not robust when faced with under-represented samples or out-of-distribution samples, which inevitably exist in real-world deployment scenarios. This paper proposes to address such challenges by exploiting semi-supervised pre-training and objective regularization to approximate subjective evaluation criteria. In particular, normalized mutual information is used to quantify the speech characteristics from the learner and the reference. An anchor model is trained using pseudo labels to predict the correctness of pronunciation. An interpolated loss function is proposed to minimize not only the prediction error with respect to ground-truth scores but also the divergence between two probability distributions estimated by the speech evaluation model and the anchor model. Compared to other state-of-the-art methods on a public data-set, this approach not only achieves high performance while evaluating the entire test-set as a whole, but also brings the most evenly distributed prediction error across distinct proficiency levels. Furthermore, empirical results show the model accuracy on out-of-distribution data also compares favorably with competitive baselines.
Authors: Yuan Liu, Shu Wang, Zhe Qu, Xingyu Li, Shichao Kan, Jianxin Wang
Abstract: Federated Domain Generalization (FedDG) aims to train the global model for generalization ability to unseen domains with multi-domain training samples. However, clients in federated learning networks are often confined to a single, non-IID domain due to inherent sampling and temporal limitations. The lack of cross-domain interaction and the in-domain divergence impede the learning of domain-common features and limit the effectiveness of existing FedDG, referred to as the single-source FedDG (sFedDG) problem. To address this, we introduce the Federated Global Consistent Augmentation (FedGCA) method, which incorporates a style-complement module to augment data samples with diverse domain styles. To ensure the effective integration of augmented samples, FedGCA employs both global guided semantic consistency and class consistency, mitigating inconsistencies from local semantics within individual clients and classes across multiple clients. The conducted extensive experiments demonstrate the superiority of FedGCA.
Authors: Hyundong Cho, Nicolaas Jedema, Leonardo F. R. Ribeiro, Karishma Sharma, Pedro Szekely, Alessandro Moschitti, Ruben Janssen, Jonathan May
Abstract: Current instruction-tuned language models are exclusively trained with textual preference data and thus are often not aligned with the unique requirements of other modalities, such as speech. To better align language models with the speech domain, we explore (i) prompting strategies grounded in radio-industry best practices and (ii) preference learning using a novel speech-based preference data of 20K samples, generated with a wide spectrum of prompts that induce varying dimensions of speech-suitability and labeled by annotators who listen to response pairs. Both human and automatic evaluation show that both prompting and preference learning increase the speech-suitability of popular instruction-tuned LLMs. Interestingly, we find that prompting and preference learning can be additive; combining them achieves the best win rates in head-to-head comparison, resulting in responses that are preferred or tied to the base model in 76.2% of comparisons on average. Lastly, we share lexical, syntactical, and qualitative analyses to showcase how each method contributes to improving the speech-suitability of generated responses.
Authors: Rui Lu, Siping Shi, Yanting Liu, Dan Wang
Abstract: As artificial intelligence continues to evolve, it is increasingly capable of handling a wide range of video analytics tasks with merely one large model. One of the key foundation technologies is the Segment Anything Model (SAM), which allows the video analytics tasks to be determined on the fly according to the input prompts from the user. However, achieving real-time response in video analytics applications is crucial for user experiences due to the limited communication and computation resources on the edge, especially with SAM, where users may continuously interact by adding or adjusting prompts. In this paper, we propose SAMEdge, a novel edge-cloud computing architecture designed to support SAM computations for edge users. SAMEdge integrates new modules on the edge and the cloud to maximize analytics accuracy under visual prompts and image prompts input with latency constraints. It addresses resource challenges associated with prompt encoding and image encoding by offering a visual prompt transformation algorithm for visual prompts and efficient workload partitioning for image encoding. SAMEdge is implemented by extending the open-source SAM project from Meta AI. We demonstrate the practical application of SAMEdge through a case study on a Visual Tour Guide application. Our evaluation indicates that SAMEdge significantly enhances the accuracy of the video analytics application under distinct network bandwidths across various prompts.
Authors: Zhanglin Wu, Daimeng Wei, Zongyao Li, Hengchao Shang, Jiaxin Guo, Shaojun Li, Zhiqiang Rao, Yuanchang Luo, Ning Xie, Hao Yang
Abstract: This paper presents the submission of Huawei Translate Services Center (HW-TSC) to the WMT24 general machine translation (MT) shared task, where we participate in the English to Chinese (en2zh) language pair. Similar to previous years' work, we use training strategies such as regularized dropout, bidirectional training, data diversification, forward translation, back translation, alternated training, curriculum learning, and transductive ensemble learning to train the neural machine translation (NMT) model based on the deep Transformer-big architecture. The difference is that we also use continue pre-training, supervised fine-tuning, and contrastive preference optimization to train the large language model (LLM) based MT model. By using Minimum Bayesian risk (MBR) decoding to select the final translation from multiple hypotheses for NMT and LLM-based MT models, our submission receives competitive results in the final evaluation.
Authors: Rakshith Jayanth, Neelesh Gupta, Viktor Prasanna
Abstract: Edge computing's growing prominence, due to its ability to reduce communication latency and enable real-time processing, is promoting the rise of high-performance, heterogeneous System-on-Chip solutions. While current approaches often involve scaling down modern hardware, the performance characteristics of neural network workloads on these platforms can vary significantly, especially when it comes to parallel processing, which is a critical consideration for edge deployments. To address this, we conduct a comprehensive study comparing the latency and throughput of various linear algebra and neural network inference tasks across CPU-only, CPU/GPU, and CPU/NPU integrated solutions. {We find that the Neural Processing Unit (NPU) excels in matrix-vector multiplication (58.6% faster) and some neural network tasks (3.2$\times$ faster for video classification and large language models). GPU outperforms in matrix multiplication (22.6% faster) and LSTM networks (2.7$\times$ faster) while CPU excels at less parallel operations like dot product. NPU-based inference offers a balance of latency and throughput at lower power consumption. GPU-based inference, though more energy-intensive, performs best with large dimensions and batch sizes. We highlight the potential of heterogeneous computing solutions for edge AI, where diverse compute units can be strategically leveraged to boost accurate and real-time inference.
Authors: Cong Wang, Zeming Chen, Shanshi Huang
Abstract: This work introduces MICSim, an open-source, pre-circuit simulator designed for early-stage evaluation of chip-level software performance and hardware overhead of mixed-signal compute-in-memory (CIM) accelerators. MICSim features a modular design, allowing easy multi-level co-design and design space exploration. Modularized from the state-of-the-art CIM simulator NeuroSim, MICSim provides a highly configurable simulation framework supporting multiple quantization algorithms, diverse circuit/architecture designs, and different memory devices. This modular approach also allows MICSim to be effectively extended to accommodate new designs. MICSim natively supports evaluating accelerators' software and hardware performance for CNNs and Transformers in Python, leveraging the popular PyTorch and HuggingFace Transformers frameworks. These capabilities make MICSim highly adaptive when simulating different networks and user-friendly. This work demonstrates that MICSim can easily be combined with optimization strategies to perform design space exploration and used for chip-level Transformers CIM accelerators evaluation. Also, MICSim can achieve a 9x - 32x speedup of NeuroSim through a statistic-based average mode proposed by this work.
Authors: John Dorsch, Maximilian Moll
Abstract: In the context of AI decision support systems (AI-DSS), we argue that meeting the demands of ethical and explainable AI (XAI) is about developing AI-DSS to provide human decision-makers with three types of human-grounded explanations: reasons, counterfactuals, and confidence, an approach we refer to as the RCC approach. We begin by reviewing current empirical XAI literature that investigates the relationship between various methods for generating model explanations (e.g., LIME, SHAP, Anchors), the perceived trustworthiness of the model, and end-user accuracy. We demonstrate how current theories about what constitutes good human-grounded reasons either do not adequately explain this evidence or do not offer sound ethical advice for development. Thus, we offer a novel theory of human-machine interaction: the theory of epistemic quasi-partnerships (EQP). Finally, we motivate adopting EQP and demonstrate how it explains the empirical evidence, offers sound ethical advice, and entails adopting the RCC approach.
Authors: Zhanglin Wu, Yuanchang Luo, Daimeng Wei, Jiawei Zheng, Bin Wei, Zongyao Li, Hengchao Shang, Jiaxin Guo, Shaojun Li, Weidong Zhang, Ning Xie, Hao Yang
Abstract: This paper presents the submission of Huawei Translation Services Center (HW-TSC) to machine translation tasks of the 20th China Conference on Machine Translation (CCMT 2024). We participate in the bilingual machine translation task and multi-domain machine translation task. For these two translation tasks, we use training strategies such as regularized dropout, bidirectional training, data diversification, forward translation, back translation, alternated training, curriculum learning, and transductive ensemble learning to train neural machine translation (NMT) models based on the deep Transformer-big architecture. Furthermore, to explore whether large language model (LLM) can help improve the translation quality of NMT systems, we use supervised fine-tuning to train llama2-13b as an Automatic post-editing (APE) model to improve the translation results of the NMT model on the multi-domain machine translation task. By using these plyometric strategies, our submission achieves a competitive result in the final evaluation.
Authors: Junyang Zhang, Mu Yuan, Ruiguang Zhong, Puhan Luo, Huiyou Zhan, Ningkang Zhang, Chengchen Hu, Xiangyang Li
Abstract: The Large Vision-Language Model (LVLM) integrates computer vision and natural language processing techniques, offering substantial application potential. However, these models demand extensive resources during inference. Adaptive attention techniques can dynamically reduce computational redundancy and thus improve efficiency. Although current adaptive attention methods significantly reduce the memory requirements of Transformer-based language models, they are not tailored for LVLMs. We observe that LVLMs generate responses from both remote image tokens and local text tokens, and different modalities have different attention patterns. This observation inspires us to manage the attention for each modality separately. Specifically, for visual input, we store the cache of potentially useful information but only compute the most critical parts. For language input, we care more about local information. Based on our observation and analysis of vision-language attention patterns, we develop A-VL, a plug-and-play adaptive attention tailored for LVLM inference. Extensive evaluations on three vision-language tasks and five datasets show the effectiveness of our designs. Our approach A-VL outperforms existing adaptive attention methods in reducing memory usage and computational load without compromising performance.
Authors: Pengfei Wang, Huanran Zheng, Silong Dai, Wenjing Yue, Wei Zhu, Xiaoling Wang
Abstract: In recent years, large language models (LLMs) have shown great potential in time-series analysis by capturing complex dependencies and improving predictive performance. However, existing approaches often struggle with modality alignment, leading to suboptimal results. To address these challenges, we present a novel framework, TS-TCD, which introduces a comprehensive three-tiered cross-modal knowledge distillation mechanism. Unlike prior work that focuses on isolated alignment techniques, our framework systematically integrates: 1) Dynamic Adaptive Gating for Input Encoding and Alignment}, ensuring coherent alignment between time-series tokens and QR-decomposed textual embeddings; 2) Layer-Wise Contrastive Learning}, aligning intermediate representations across modalities to reduce feature-level discrepancies; and 3) Optimal Transport-Driven Output Alignment}, which ensures consistent output predictions through fine-grained cross-modal alignment. Extensive experiments on benchmark time-series datasets demonstrate that TS-TCD achieves state-of-the-art results, outperforming traditional methods in both accuracy and robustness.
Authors: Hong Chen, Xin Wang, Yuwei Zhou, Bin Huang, Yipeng Zhang, Wei Feng, Houlun Chen, Zeyang Zhang, Siao Tang, Wenwu Zhu
Abstract: Multi-modal generative AI has received increasing attention in both academia and industry. Particularly, two dominant families of techniques are: i) The multi-modal large language model (MLLM) such as GPT-4V, which shows impressive ability for multi-modal understanding; ii) The diffusion model such as Sora, which exhibits remarkable multi-modal powers, especially with respect to visual generation. As such, one natural question arises: Is it possible to have a unified model for both understanding and generation? To answer this question, in this paper, we first provide a detailed review of both MLLM and diffusion models, including their probabilistic modeling procedure, multi-modal architecture design, and advanced applications to image/video large language models as well as text-to-image/video generation. Then, we discuss the two important questions on the unified model: i) whether the unified model should adopt the auto-regressive or diffusion probabilistic modeling, and ii) whether the model should utilize a dense architecture or the Mixture of Experts(MoE) architectures to better support generation and understanding, two objectives. We further provide several possible strategies for building a unified model and analyze their potential advantages and disadvantages. We also summarize existing large-scale multi-modal datasets for better model pretraining in the future. To conclude the paper, we present several challenging future directions, which we believe can contribute to the ongoing advancement of multi-modal generative AI.
Authors: Furkan Pala, Mehmet Yasin Akp{\i}nar, Onur Deniz, G\"ul\c{s}en Eryi\u{g}it
Abstract: Multimodal key information extraction (KIE) models have been studied extensively on semi-structured documents. However, their investigation on unstructured documents is an emerging research topic. The paper presents an approach to adapt a multimodal transformer (i.e., ViBERTgrid previously explored on semi-structured documents) for unstructured financial documents, by incorporating a BiLSTM-CRF layer. The proposed ViBERTgrid BiLSTM-CRF model demonstrates a significant improvement in performance (up to 2 percentage points) on named entity recognition from unstructured documents in financial domain, while maintaining its KIE performance on semi-structured documents. As an additional contribution, we publicly released token-level annotations for the SROIE dataset in order to pave the way for its use in multimodal sequence labeling models.
Authors: Hansa Mohanty, Deepak Narayan Gadde
Abstract: The complexity of modern-day System-on-Chips (SoCs) is continually increasing, and it becomes increasingly challenging to deliver dependable and credible chips in a short time-to-market. Especially, in the case of test chips, where the aim is to study the feasibility of the design, time is a crucial factor. Pre-silicon functional verification is one of the main contributors that makes up a large portion of the product development cycle. Verification engineers often loosely verify test chips that turn out to be non-functional on the silicon, ultimately resulting in expensive re-spins. To left-shift the verification efforts, formal verification is a powerful methodology that aims to exhaustively verify designs, giving better confidence in the overall quality. This paper focuses on the pragmatic formal verification of a mixed signal Intellectual Property (IP) that has a combination of digital and analog blocks. This paper discusses a novel approach of including the analog behavioral model into the formal verification setup. Digital and Analog Mixed-Signal (AMS) designs, which are fundamentally different in nature, are integrated seamlessly in a formal verification setup, a concept that can be referred to as "Analogous Alignments". Our formal setup leverages powerful formal techniques such as FPV, CSR verification, and connectivity checks. The properties used for FPV are auto-generated using a metamodeling framework. The paper also discusses the challenges faced especially related to state-space explosion, non-compatibility of formal with AMS models, and techniques to mitigate them such as k-induction. With this verification approach, we were able to exhaustively verify the design within a reasonable time and with sufficient coverage. We also reported several bugs at an early stage, making the complete design verification process iterative and effective.
Authors: Kevin Baum, Lisa Dargasz, Felix Jahn, Timo P. Gros, Verena Wolf
Abstract: We propose an extension of the reinforcement learning architecture that enables moral decision-making of reinforcement learning agents based on normative reasons. Central to this approach is a reason-based shield generator yielding a moral shield that binds the agent to actions that conform with recognized normative reasons so that our overall architecture restricts the agent to actions that are (internally) morally justified. In addition, we describe an algorithm that allows to iteratively improve the reason-based shield generator through case-based feedback from a moral judge.
Authors: Ye Han, Lijun Zhang, Dejian Meng, Xingyu Hu, Yixia Lu
Abstract: In mixed autonomy traffic environment, every decision made by an autonomous-driving car may have a great impact on the transportation system. Because of the complex interaction between vehicles, it is challenging to make decisions that can ensure both high traffic efficiency and safety now and futher. Connected automated vehicles (CAVs) have great potential to improve the quality of decision-making in this continuous, highly dynamic and interactive environment because of their stronger sensing and communicating ability. For multi-vehicle collaborative decision-making algorithms based on deep reinforcement learning (DRL), we need to represent the interactions between vehicles to obtain interactive features. The representation in this aspect directly affects the learning efficiency and the quality of the learned policy. To this end, we propose a CAV decision-making architecture based on transformer and reinforcement learning algorithms. A learnable policy token is used as the learning medium of the multi-vehicle joint policy, the states of all vehicles in the area of interest can be adaptively noticed in order to extract interactive features among agents. We also design an intuitive physical positional encodings, the redundant location information of which optimizes the performance of the network. Simulations show that our model can make good use of all the state information of vehicles in traffic scenario, so as to obtain high-quality driving decisions that meet efficiency and safety objectives. The comparison shows that our method significantly improves existing DRL-based multi-vehicle cooperative decision-making algorithms.
Authors: Pablo Fern\'andez-Saborido, Marcos Fern\'andez-Pichel, David E. Losada
Abstract: Evaluating the capabilities of Large Language Models (LLMs) to assist teachers and students in educational tasks is receiving increasing attention. In this paper, we assess ChatGPT's capacities to solve and grade real programming exams, from an accredited BSc degree in Computer Science, written in Spanish. Our findings suggest that this AI model is only effective for solving simple coding tasks. Its proficiency in tackling complex problems or evaluating solutions authored by others are far from effective. As part of this research, we also release a new corpus of programming tasks and the corresponding prompts for solving the problems or grading the solutions. This resource can be further exploited by other research teams.
Authors: Lucas Heublein, Tobias Feigl, Thorsten Nowak, Alexander R\"ugamer, Christopher Mutschler, Felix Ott
Abstract: Jamming devices present a significant threat by disrupting signals from the global navigation satellite system (GNSS), compromising the robustness of accurate positioning. The detection of anomalies within frequency snapshots is crucial to counteract these interferences effectively. A critical preliminary measure involves the reliable classification of interferences and characterization and localization of jamming devices. This paper introduces an extensive dataset compromising snapshots obtained from a low-frequency antenna, capturing diverse generated interferences within a large-scale environment including controlled multipath effects. Our objective is to assess the resilience of ML models against environmental changes, such as multipath effects, variations in interference attributes, such as the interference class, bandwidth, and signal-to-noise ratio, the accuracy jamming device localization, and the constraints imposed by snapshot input lengths. By analyzing the aleatoric and epistemic uncertainties, we demonstrate the adaptness of our model in generalizing across diverse facets, thus establishing its suitability for real-world applications. https://gitlab.cc-asp.fraunhofer.de/darcy_gnss/controlled_low_frequency
URLs: https://gitlab.cc-asp.fraunhofer.de/darcy_gnss/controlled_low_frequency
Authors: Ismail Labiad, Thomas B\"ack, Pierre Fernandez, Laurent Najman, Tom Sanders, Furong Ye, Mariia Zameshina, Olivier Teytaud
Abstract: In many cases, adversarial attacks are based on specialized algorithms specifically dedicated to attacking automatic image classifiers. These algorithms perform well, thanks to an excellent ad hoc distribution of initial attacks. However, these attacks are easily detected due to their specific initial distribution. We therefore consider other black-box attacks, inspired from generic black-box optimization tools, and in particular the log-normal algorithm. We apply the log-normal method to the attack of fake detectors, and get successful attacks: importantly, these attacks are not detected by detectors specialized on classical adversarial attacks. Then, combining these attacks and deep detection, we create improved fake detectors.
Authors: Jordi Bayarri-Planas, Ashwin Kumar Gururajan, Dario Garcia-Gasulla
Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities in natural language processing, and yet, their factual inaccuracies and hallucinations limits their application, particularly in critical domains like healthcare. Context retrieval methods, by introducing relevant information as input, have emerged as a crucial approach for enhancing LLM factuality and reliability. This study explores the boundaries of context retrieval methods within the healthcare domain, optimizing their components and benchmarking their performance against open and closed alternatives. Our findings reveal how open LLMs, when augmented with an optimized retrieval system, can achieve performance comparable to the biggest private solutions on established healthcare benchmarks (multiple-choice question answering). Recognizing the lack of realism of including the possible answers within the question (a setup only found in medical exams), and after assessing a strong LLM performance degradation in the absence of those options, we extend the context retrieval system in that direction. In particular, we propose OpenMedPrompt a pipeline that improves the generation of more reliable open-ended answers, moving this technology closer to practical application.
Authors: Alessio Pellegrino, \"Ozg\"ur Akg\"un, Nguyen Dang, Zeynep Kiziltan, Ian Miguel
Abstract: Constraint modelling languages such as Essence offer a means to describe combinatorial problems at a high-level, i.e., without committing to detailed modelling decisions for a particular solver or solving paradigm. Given a problem description written in Essence, there are multiple ways to translate it to a low-level constraint model. Choosing the right combination of a low-level constraint model and a target constraint solver can have significant impact on the effectiveness of the solving process. Furthermore, the choice of the best combination of constraint model and solver can be instance-dependent, i.e., there may not exist a single combination that works best for all instances of the same problem. In this paper, we consider the task of building machine learning models to automatically select the best combination for a problem instance. A critical part of the learning process is to define instance features, which serve as input to the selection model. Our contribution is automatic learning of instance features directly from the high-level representation of a problem instance using a language model. We evaluate the performance of our approach using the Essence modelling language with a case study involving the car sequencing problem.
Authors: Rui Gan, Haotian Shi, Pei Li, Keshu Wu, Bocheng An, Linheng Li, Junyi Ma, Chengyuan Ma, Bin Ran
Abstract: Vehicle trajectory prediction plays a vital role in intelligent transportation systems and autonomous driving, as it significantly affects vehicle behavior planning and control, thereby influencing traffic safety and efficiency. Numerous studies have been conducted to predict short-term vehicle trajectories in the immediate future. However, long-term trajectory prediction remains a major challenge due to accumulated errors and uncertainties. Additionally, balancing accuracy with interpretability in the prediction is another challenging issue in predicting vehicle trajectory. To address these challenges, this paper proposes a Goal-based Neural Physics Vehicle Trajectory Prediction Model (GNP). The GNP model simplifies vehicle trajectory prediction into a two-stage process: determining the vehicle's goal and then choosing the appropriate trajectory to reach this goal. The GNP model contains two sub-modules to achieve this process. The first sub-module employs a multi-head attention mechanism to accurately predict goals. The second sub-module integrates a deep learning model with a physics-based social force model to progressively predict the complete trajectory using the generated goals. The GNP demonstrates state-of-the-art long-term prediction accuracy compared to four baseline models. We provide interpretable visualization results to highlight the multi-modality and inherent nature of our neural physics framework. Additionally, ablation studies are performed to validate the effectiveness of our key designs.
Authors: Jonathan Paul Driemeyer Brown, Tiago Oliveira Weber
Abstract: The usefulness of Large Language Models (LLM) is being continuously tested in various fields. However, their intrinsic linguistic characteristic is still one of the limiting factors when applying these models to exact sciences. In this article, a novel approach to use General Pre-Trained Transformers to assist in the design phase of data acquisition systems will be presented. The solution is packaged in the form of an application that retains the conversational aspects of LLMs, in such a manner that the user must provide details on the desired project in order for the model to draft both a system-level architectural diagram and the block-level specifications, following a Top-Down methodology based on restrictions. To test this tool, two distinct user emulations were used, one of which uses an additional GPT model. In total, 4 different data acquisition projects were used in the testing phase, each with its own measurement requirements: angular position, temperature, acceleration and a fourth project with both pressure and superficial temperature measurements. After 160 test iterations, the study concludes that there is potential for these models to serve adequately as synthesis/assistant tools for data acquisition systems, but there are still technological limitations. The results show coherent architectures and topologies, but that GPTs have difficulties in simultaneously considering all requirements and many times commits theoretical mistakes.
Authors: Truong Thanh Hung Nguyen, Phuc Truong Loc Nguyen, Monica Wachowicz, Hung Cao
Abstract: This paper presents a Multimodal Ambient Context-enriched Intelligence Platform (MACeIP) for Smart Cities, a comprehensive system designed to enhance urban management and citizen engagement. Our platform integrates advanced technologies, including Internet of Things (IoT) sensors, edge and cloud computing, and Multimodal AI, to create a responsive and intelligent urban ecosystem. Key components include Interactive Hubs for citizen interaction, an extensive IoT sensor network, intelligent public asset management, a pedestrian monitoring system, a City Planning Portal, and a Cloud Computing System. We demonstrate the prototype of MACeIP in several cities, focusing on Fredericton, New Brunswick. This work contributes to innovative city development by offering a scalable, efficient, and user-centric approach to urban intelligence and management.
Authors: Yi-Fei Zhao, Allyn Bove, David Thompson, James Hill, Yi Xu, Yufan Ren, Andrea Hassman, Leming Zhou, Yanshan Wang
Abstract: Low back pain (LBP) is a leading cause of disability globally. Following the onset of LBP and subsequent treatment, adequate patient education is crucial for improving functionality and long-term outcomes. Despite advancements in patient education strategies, significant gaps persist in delivering personalized, evidence-based information to patients with LBP. Recent advancements in large language models (LLMs) and generative artificial intelligence (GenAI) have demonstrated the potential to enhance patient education. However, their application and efficacy in delivering educational content to patients with LBP remain underexplored and warrant further investigation. In this study, we introduce a novel approach utilizing LLMs with Retrieval-Augmented Generation (RAG) and few-shot learning to generate tailored educational materials for patients with LBP. Physical therapists manually evaluated our model responses for redundancy, accuracy, and completeness using a Likert scale. In addition, the readability of the generated education materials is assessed using the Flesch Reading Ease score. The findings demonstrate that RAG-based LLMs outperform traditional LLMs, providing more accurate, complete, and readable patient education materials with less redundancy. Having said that, our analysis reveals that the generated materials are not yet ready for use in clinical practice. This study underscores the potential of AI-driven models utilizing RAG to improve patient education for LBP; however, significant challenges remain in ensuring the clinical relevance and granularity of content generated by these models.
Authors: Thierry Petit, Arnault Pachot, Claire Conan-Vrinat, Alexandre Dubarry
Abstract: This article introduces an innovative architecture designed to declaratively combine Large Language Models (LLMs) with shared histories, and triggers to identify the most appropriate LLM for a given task. Our approach is general and declarative, relying on the construction of finite automata coupled with an event management system. The developed tool is crafted to facilitate the efficient and complex integration of LLMs with minimal programming effort, especially, but not only, for integrating methods of positive psychology to AI. The flexibility of our technique is demonstrated through applied examples in automation, communication, and ethics.
Authors: Shuo Yu (Anhui Province Key Laboratory of Big Data Analysis and Application, University of Science and Technology of China, State Key Laboratory of Cognitive Intelligence), Mingyue Cheng (Anhui Province Key Laboratory of Big Data Analysis and Application, University of Science and Technology of China, State Key Laboratory of Cognitive Intelligence), Jiqian Yang (Anhui Province Key Laboratory of Big Data Analysis and Application, University of Science and Technology of China, State Key Laboratory of Cognitive Intelligence), Jie Ouyang (Anhui Province Key Laboratory of Big Data Analysis and Application, University of Science and Technology of China, State Key Laboratory of Cognitive Intelligence)
Abstract: Retrieval-Augmented Generation (RAG) enhances generative models by integrating retrieval mechanisms, which allow these models to access and utilize external knowledge sources. Despite its advantages, RAG encounters significant challenges, particularly in effectively handling real-world queries and mitigating hallucinations. The KDD Cup 2024 CRAG competition brings these issues to the forefront by incorporating both web pages and a mock API as knowledge sources, adding the complexity of parsing HTML before large language models (LLMs) can process the information. In this paper, we propose a novel RAG benchmark designed to address these challenges. Our work provides a comprehensive set of experimental results, offering valuable insights for the study of RAG. We thoroughly examine the entire RAG process, including knowledge source selection, retrieval, organization, and reasoning. Key findings from our study include the impact of automated knowledge source selection using agents and the influence of noise chunks on RAG reasoning. Additionally, we conduct detailed experiments to analyze the effects of various hyperparameters on RAG performance. To support further research, we have made our results, the associated code, and a parsed version of the CRAG dataset publicly available\footnote{https://github.com/USTCAGI/RAG-X}, contributing to the advancement of RAG methodologies and establishing a solid foundation for future work in this domain.
Authors: Yun Joon Soh, Hanxian Huang, Yuandong Tian, Jishen Zhao
Abstract: Supporting longer context for Large Language Models (LLM) is a promising direction to advance LLMs. As training a model for a longer context window is computationally expensive, many alternative solutions, such as Retrieval Augmented Generation (RAG), have been used. However, most existing RAG methods adopt embedding-based retrieval that falls short on long contexts. To address such challenges, we propose an attention-based retrieval technique, You Only Use Reactive Attention slice (YOURA). YOURA leverages a novel retrieval heuristic called reaction score to rank the relevance of each sentence in the input context with the query sentence. Intuitively, we measure how the per-token attention score "reacts" to the query and greedily retrieves the most reactive sentences. Internally, YOURA generates a token-indexed vector (called reaction vector) for the whole input context. To map each sentence to the token-indexed vector, we propose an Embedding-Agnostic Sentence Yield (EASY), a best-effort token wiggling algorithm. We evaluate our retrieval technique on three open-source pre-trained LLM models across six LongBench QA datasets. Our technique achieves up to 30% vLLM inference throughput improvement for serving long-context queries with a nearly identical quality score to the simple yet effective truncate-middle approach.
Authors: Aman Bhargava, Cameron Witkowski, Alexander Detkov, Matt Thomson
Abstract: Two primary ways to change LLM behavior are prompting and weight updates (e.g., fine-tuning). Prompting LLMs is simple and effective, specifying the desired changes explicitly in natural language, whereas weight updates provide more expressive and permanent behavior changes, specified implicitly via training on large datasets. We present a technique for "baking" prompts into the weights of an LLM. Prompt Baking converts a prompt $u$ and initial weights $\theta$ to a new set of weights $\theta_u$ such that new "baked" LLM behaves like the original prompted LLM. Mathematically, we minimize the KL divergence between $P_\theta(\cdot | u)$ and $P_{\theta_u}(\cdot)$, where $P$ is the LLM's probability distribution over token sequences. Across all our experiments, we find prompts can be readily baked into weight updates. Baking chain-of-thought prompts improves zero-shot performance on GSM8K, ASDiv, MBPP, ARC-Easy, ARC-Challenge, and CommonsenseQA benchmarks. Baking news headlines directly updates an LLM's knowledge. And baking instructions & personas alleviates "prompt forgetting" over long sequences. Furthermore, stopping baking early creates "half-baked" models, continuously scaling prompt strength. Baked models retain their sensitivity to further prompting and baking, including re-prompting with the baked-in prompt. Surprisingly, the re-prompted models yield further performance gains in instruction following, as well as math reasoning and coding benchmarks. Taking re-prompting and re-baking to the limit yields a form of iterative self-improvement we call Prompt Pursuit, and preliminary results on instruction following exhibit dramatic performance gains. Finally, we discuss implications for AI safety, continuous model updating, enhancing real-time learning capabilities in LLM-based agents, and generating more stable AI personas.
Authors: Yuqian Wu, Yuhong Peng, Jiapeng Yu, Raymond S. T. Lee
Abstract: LLM-based Multi-Agent Systems have potential benefits of complex decision-making tasks management across various domains but their applications in the next Point-of-Interest (POI) recommendation remain underexplored. This paper proposes a novel MAS4POI system designed to enhance next POI recommendations through multi-agent interactions. MAS4POI supports Large Language Models (LLMs) specializing in distinct agents such as DataAgent, Manager, Analyst, and Navigator with each contributes to a collaborative process of generating the next POI recommendations.The system is examined by integrating six distinct LLMs and evaluated by two real-world datasets for recommendation accuracy improvement in real-world scenarios. Our code is available at https://github.com/yuqian2003/MAS4POI.
Authors: Minghao Liu, Mingxiu Sui, Cangqing Wang, Zhejie Zhou
Abstract: Effective communication in automated chat systems hinges on the ability to understand and respond to context. Traditional models often struggle with determining when additional context is necessary for generating appropriate responses. This paper introduces Context-Aware BERT (CA-BERT), a transformer-based model specifically fine-tuned to address this challenge. CA-BERT innovatively applies deep learning techniques to discern context necessity in multi-turn chat interactions, enhancing both the relevance and accuracy of responses. We describe the development of CA-BERT, which adapts the robust architecture of BERT with a novel training regimen focused on a specialized dataset of chat dialogues. The model is evaluated on its ability to classify context necessity, demonstrating superior performance over baseline BERT models in terms of accuracy and efficiency. Furthermore, CA-BERT's implementation showcases significant reductions in training time and resource usage, making it feasible for real-time applications. The results indicate that CA-BERT can effectively enhance the functionality of chatbots by providing a nuanced understanding of context, thereby improving user experience and interaction quality in automated systems. This study not only advances the field of NLP in chat applications but also provides a framework for future research into context-sensitive AI developments.
Authors: Olivia Sturman, Aparna Joshi, Bhaktipriya Radharapu, Piyush Kumar, Renee Shelby
Abstract: Increasing use of large language models (LLMs) demand performant guardrails to ensure the safety of inputs and outputs of LLMs. When these safeguards are trained on imbalanced data, they can learn the societal biases. We present a light-weight, post-processing method for mitigating counterfactual fairness in closed-source text safety classifiers. Our approach involves building an ensemble that not only outperforms the input classifiers and policy-aligns them, but also acts as a debiasing regularizer. We introduce two threshold-agnostic metrics to assess the counterfactual fairness of a model, and demonstrate how combining these metrics with Fair Data Reweighting (FDW) helps mitigate biases. We create an expanded Open AI dataset, and a new templated LLM-generated dataset based on user-prompts, both of which are counterfactually balanced across identity groups and cover four key areas of safety; we will work towards publicly releasing these datasets. Our results show that our approach improves counterfactual fairness with minimal impact on model performance.
Authors: Paulina Toro Isaza, Michael Nidd, Noah Zheutlin, Jae-wook Ahn, Chidansh Amitkumar Bhatt, Yu Deng, Ruchi Mahindru, Martin Franz, Hans Florian, Salim Roukos
Abstract: Clients wishing to implement generative AI in the domain of IT Support and AIOps face two critical issues: domain coverage and model size constraints due to model choice limitations. Clients might choose to not use larger proprietary models such as GPT-4 due to cost and privacy concerns and so are limited to smaller models with potentially less domain coverage that do not generalize to the client's domain. Retrieval augmented generation is a common solution that addresses both of these issues: a retrieval system first retrieves the necessary domain knowledge which a smaller generative model leverages as context for generation. We present a system developed for a client in the IT Support domain for support case solution recommendation that combines retrieval augmented generation (RAG) for answer generation with an encoder-only model for classification and a generative large language model for query generation. We cover architecture details, data collection and annotation, development journey and preliminary validations, expected final deployment process and evaluation plans, and finally lessons learned.
Authors: Art\=urs Kanepajs, Vladimir Ivanov, Richard Moulange
Abstract: Linguistically inclusive LLMs -- which maintain good performance regardless of the language with which they are prompted -- are necessary for the diffusion of AI benefits around the world. Multilingual jailbreaks that rely on language translation to evade safety measures undermine the safe and inclusive deployment of AI systems. We provide policy recommendations to enhance the multilingual capabilities of AI while mitigating the risks of multilingual jailbreaks. We quantitatively assess the relationship between language resourcedness and model vulnerabilities to multilingual jailbreaks for five frontier large language models across 24 official EU languages. Building on prior research, we propose policy actions that align with the EU legal landscape and institutional framework to address multilingual jailbreaks, while promoting linguistic inclusivity. These include mandatory assessments of multilingual capabilities and vulnerabilities, public opinion research, and state support for multilingual AI development. The measures aim to improve AI safety and functionality through EU policy initiatives, guiding the implementation of the EU AI Act and informing regulatory efforts of the European AI Office.
Authors: Margherita Martorana, Xueli Pan, Benno Kruit, Tobias Kuhn, Jacco van Ossenbruggen
Abstract: Traditional Semantic Table Interpretation (STI) methods rely primarily on the underlying table data to create semantic annotations. This year's SemTab challenge introduced the ``Metadata to KG'' track, which focuses on performing STI by using only metadata information, without access to the underlying data. In response to this new challenge, we introduce a new term: Column Vocabulary Association (CVA). This term refers to the task of semantic annotation of column headers solely based on metadata information. In this study, we evaluate the performance of various methods in executing the CVA task, including a Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) approach, as well as a more traditional similarity approach with SemanticBERT. Our methodology uses a zero-shot setting, with no pretraining or examples passed to the Large Language Models (LLMs), as we aim to avoid a domain-specific setting. We investigate a total of 7 different LLMs, of which three commercial GPT models (i.e. gpt-3.5-turbo-0.125, gpt-4o and gpt-4-turbo) and four open source models (i.e. llama3-80b, llama3-7b, gemma-7b and mixtral-8x7b). We integrate this models with RAG systems, and we explore how variations in temperature settings affect performances. Moreover, we continue our investigation by performing the CVA task utilizing SemanticBERT, analyzing how various metadata information influence its performance. Initial findings indicate that LLMs generally perform well at temperatures below 1.0, achieving an accuracy of 100\% in certain cases. Nevertheless, our investigation also reveal that the nature of the data significantly influences CVA task outcomes. In fact, in cases where the input data and glossary are related (for example by being created by the same organizations) traditional methods appear to surpass the performance of LLMs.
Authors: Maria Wang, Srinivas Sunkara, Gilles Baechler, Jason Lin, Yun Zhu, Fedir Zubach, Lei Shu, Jindong Chen
Abstract: The rise of multimodal LLMs and web agents calls for the creation of challenging benchmarks to evaluate neural architectures. Unlike existing benchmarks that focus on multi-step web navigation, we present WebQuest, a multi-page question-answering dataset that requires simultaneous retrieval and reasoning across web interaction sequences grounded in real-world usage. WebQuest includes three question categories: single-screen reasoning, multi-screen reasoning, and questions based on navigation traces. We evaluate some of the leading multimodal models like GPT-4V, Gemini Flash, and Claude 3 on our dataset, revealing a significant gap between single-screen and multi-screen reasoning. Finally, we investigate inference time techniques like Chain-of-Thought prompting to improve model capabilities on multi-screen reasoning.
Authors: Yi Xu, Bo Xue, Shuqian Sheng, Cheng Deng, Jiaxin Ding, Zanwei Shen, Luoyi Fu, Xinbing Wang, Chenghu Zhou
Abstract: In the ever-expanding landscape of academic research, the proliferation of ideas presents a significant challenge for researchers: discerning valuable ideas from the less impactful ones. The ability to efficiently evaluate the potential of these ideas is crucial for the advancement of science and paper review. In this work, we focus on idea assessment, which aims to leverage the knowledge of large language models to assess the merit of scientific ideas. First, we investigate existing text evaluation research and define the problem of quantitative evaluation of ideas. Second, we curate and release a benchmark dataset from nearly four thousand manuscript papers with full texts, meticulously designed to train and evaluate the performance of different approaches to this task. Third, we establish a framework for quantifying the value of ideas by employing representations in a specific layer of large language models. Experimental results show that the scores predicted by our method are relatively consistent with those of humans. Our findings suggest that the representations of large language models hold more potential in quantifying the value of ideas than their generative outputs, demonstrating a promising avenue for automating the idea assessment process.
Authors: Hannes Thurnherr, J\'er\'emy Scheurer
Abstract: Achieving a mechanistic understanding of transformer-based language models is an open challenge, especially due to their large number of parameters. Moreover, the lack of ground truth mappings between model weights and their functional roles hinders the effective evaluation of interpretability methods, impeding overall progress. Tracr, a method for generating compiled transformers with inherent ground truth mappings in RASP, has been proposed to address this issue. However, manually creating a large number of models needed for verifying interpretability methods is labour-intensive and time-consuming. In this work, we present a novel approach for generating interpretability test beds using large language models (LLMs) and introduce TracrBench, a novel dataset consisting of 121 manually written and LLM-generated, human-validated RASP programs and their corresponding transformer weights. During this process, we evaluate the ability of frontier LLMs to autonomously generate RASP programs and find that this task poses significant challenges. GPT-4-turbo, with a 20-shot prompt and best-of-5 sampling, correctly implements only 57 out of 101 test programs, necessitating the manual implementation of the remaining programs. With its 121 samples, TracrBench aims to serve as a valuable testbed for evaluating and comparing interpretability methods.
Authors: Maria Tsfasman, Bernd Dudzik, Kristian Fenech, Andras Lorincz, Catholijn M. Jonker, Catharine Oertel
Abstract: The quality of human social relationships is intricately linked to human memory processes, with memory serving as the foundation for the creation of social bonds. Since human memory is selective, differing recollections of the same events within a group can lead to misunderstandings and misalignments in what is perceived to be common ground in the group. Yet, conversational facilitation systems, aimed at advancing the quality of group interactions, usually focus on tracking users' states within an individual session, ignoring what remains in each participant's memory after the interaction. Conversational memory is the process by which humans encode, retain and retrieve verbal, non-verbal and contextual information from a conversation. Understanding conversational memory can be used as a source of information on the long-term development of social connections within a group. This paper introduces the MeMo corpus, the first conversational dataset annotated with participants' memory retention reports, aimed at facilitating computational modelling of human conversational memory. The MeMo corpus includes 31 hours of small-group discussions on the topic of Covid-19, repeated over the term of 2 weeks. It integrates validated behavioural and perceptual measures, and includes audio, video, and multimodal annotations, offering a valuable resource for studying and modelling conversational memory and group dynamics. By introducing the MeMo corpus, presenting an analysis of its validity, and demonstrating its usefulness for future research, this paper aims to pave the way for future research in conversational memory modelling for intelligent system development.
Authors: Yiheng Wu, Roman Yangarber, Xian Mao
Abstract: The remarkable capabilities of Large Language Models (LLMs) in text comprehension and generation have revolutionized Information Extraction (IE). One such advancement is in Document-level Relation Triplet Extraction (DocRTE), a critical task in information systems that aims to extract entities and their semantic relationships from documents. However, existing methods are primarily designed for Sentence level Relation Triplet Extraction (SentRTE), which typically handles a limited set of relations and triplet facts within a single sentence. Additionally, some approaches treat relations as candidate choices integrated into prompt templates, resulting in inefficient processing and suboptimal performance when determining the relation elements in triplets. To address these limitations, we introduce a Discriminative and Voice Aware Paradigm DiVA. DiVA involves only two steps: performing document-level relation extraction (DocRE) and then identifying the subject object entities based on the relation. No additional processing is required simply input the document to directly obtain the triplets. This streamlined process more accurately reflects real-world scenarios for triplet extraction. Our innovation lies in transforming DocRE into a discriminative task, where the model pays attention to each relation and to the often overlooked issue of active vs. passive voice within the triplet. Our experiments on the Re-DocRED and DocRED datasets demonstrate state-of-the-art results for the DocRTE task.
Authors: Harikha Manthena, Shaghayegh Shajarian, Jeffrey Kimmell, Mahmoud Abdelsalam, Sajad Khorsandroo, Maanak Gupta
Abstract: Machine learning (ML) has seen exponential growth in recent years, finding applications in various domains such as finance, medicine, and cybersecurity. Malware remains a significant threat to modern computing, frequently used by attackers to compromise systems. While numerous machine learning-based approaches for malware detection achieve high performance, they often lack transparency and fail to explain their predictions. This is a critical drawback in malware analysis, where understanding the rationale behind detections is essential for security analysts to verify and disseminate information. Explainable AI (XAI) addresses this issue by maintaining high accuracy while producing models that provide clear, understandable explanations for their decisions. In this survey, we comprehensively review the current state-of-the-art ML-based malware detection techniques and popular XAI approaches. Additionally, we discuss research implementations and the challenges of explainable malware analysis. This theoretical survey serves as an entry point for researchers interested in XAI applications in malware detection. By analyzing recent advancements in explainable malware analysis, we offer a broad overview of the progress in this field, positioning our work as the first to extensively cover XAI methods for malware classification and detection.
Authors: Diego Calanzone, Stefano Teso, Antonio Vergari
Abstract: Large language models (LLMs) are a promising venue for natural language understanding and generation. However, current LLMs are far from reliable: they are prone to generating non-factual information and, more crucially, to contradicting themselves when prompted to reason about relations between entities of the world. These problems are currently addressed with large scale fine-tuning or by delegating reasoning to external tools. In this work, we strive for a middle ground and introduce a loss based on neuro-symbolic reasoning that teaches an LLM to be logically consistent with an external set of facts and rules and improves self-consistency even when the LLM is fine-tuned on a limited set of facts. Our approach also allows to easily combine multiple logical constraints at once in a principled way, delivering LLMs that are more consistent w.r.t. all constraints and improve over several baselines w.r.t. a given constraint. Moreover, our method allows LLMs to extrapolate to unseen but semantically similar factual knowledge, represented in unseen datasets, more systematically.
Authors: Marius Funk, Shogo Okada, Elisabeth Andr\'e
Abstract: Non-verbal behavior is a central challenge in understanding the dynamics of a conversation and the affective states between interlocutors arising from the interaction. Although psychological research has demonstrated that non-verbal behaviors vary across cultures, limited computational analysis has been conducted to clarify these differences and assess their impact on engagement recognition. To gain a greater understanding of engagement and non-verbal behaviors among a wide range of cultures and language spheres, in this study we conduct a multilingual computational analysis of non-verbal features and investigate their role in engagement and engagement prediction. To achieve this goal, we first expanded the NoXi dataset, which contains interaction data from participants living in France, Germany, and the United Kingdom, by collecting session data of dyadic conversations in Japanese and Chinese, resulting in the enhanced dataset NoXi+J. Next, we extracted multimodal non-verbal features, including speech acoustics, facial expressions, backchanneling and gestures, via various pattern recognition techniques and algorithms. Then, we conducted a statistical analysis of listening behaviors and backchannel patterns to identify culturally dependent and independent features in each language and common features among multiple languages. These features were also correlated with the engagement shown by the interlocutors. Finally, we analyzed the influence of cultural differences in the input features of LSTM models trained to predict engagement for five language datasets. A SHAP analysis combined with transfer learning confirmed a considerable correlation between the importance of input features for a language set and the significant cultural characteristics analyzed.
Authors: Zhen Yang, Jinhao Chen, Zhengxiao Du, Wenmeng Yu, Weihan Wang, Wenyi Hong, Zhihuan Jiang, Bin Xu, Yuxiao Dong, Jie Tang
Abstract: Large language models (LLMs) have demonstrated significant capabilities in mathematical reasoning, particularly with text-based mathematical problems. However, current multi-modal large language models (MLLMs), especially those specialized in mathematics, tend to focus predominantly on solving geometric problems but ignore the diversity of visual information available in other areas of mathematics. Moreover, the geometric information for these specialized mathematical MLLMs is derived from several public datasets, which are typically limited in diversity and complexity. To address these limitations, we aim to construct a fine-tuning dataset named MathVL, and develop a series of specialized mathematical MLLMs termed MathGLM-Vision by conducting Supervised Fine-Tuning (SFT) on MathVL with various parameter-scale backbones. To extensively evaluate the effectiveness of MathGLM-Vision, we conduct experiments on several public benchmarks and our curated MathVL-test consisting of 2,000 problems. Experimental results demonstrate that MathGLM-Vision achieves significant improvements compared with some existing models, including backbone models and open-source mathematical MLLMs. These findings indicate the importance of diversity dataset in enhancing the mathematical reasoning abilities of MLLMs.
Authors: Lei Liang, Mengshu Sun, Zhengke Gui, Zhongshu Zhu, Zhouyu Jiang, Ling Zhong, Yuan Qu, Peilong Zhao, Zhongpu Bo, Jin Yang, Huaidong Xiong, Lin Yuan, Jun Xu, Zaoyang Wang, Wen Zhang, Huajun Chen, Zhiqiang Zhang, Jun Zhou
Abstract: The recently developed retrieval-augmented generation (RAG) technology enables the efficient construction of domain-specific applications. However, it faces limitations due to fuzzy retrieval processes, the "hallucination" problem of understanding and reasoning capabilities of general language models, and cascading losses in complex systems. These challenges hinder the effectiveness of specialized knowledge services. However, in scenarios such as scientific computing, medicine, and law, the accuracy of knowledge, the completeness of information, and the logical rigor of rules, time, and values are particularly critical. We Introduce professional domain knowledge service framework: Knowledge Augmented Generation(KAG) to improve generation and reasoning performance by bidirectionally enhancing large language model(LLM)s and knowledge graph(KG)s, including five key enhancements: 1) LLM-friendly knowledge semantic representation, 2) mutual indexing between knowledge graph and original chunks, 3) logicalform-guided hybrid reasoning and solving, 4) Knowledge alignment based on semantic reasoning, 5) Model for KAG. We compared KAG with existing RAG methods in multi-hop question answering. The results show that KAG performs significantly better than the state-of-the-art methods, with a relative improvement from 19.6% to 33.4% in F1. We apply KAG to two professional knowledge Q&A tasks of Ant Group, including E-Goverment Q&A and E-Health Q&A, and has achieved significant improvement in professionalism compared with NaiveRAG. We will soon natively support KAG on the open source KG engine OpenSPG, allowing developers to more easily build rigorous knowledge decision-making or convenient information retrieval services.
Authors: Kuan Wang, Alexander Bukharin, Haoming Jiang, Qingyu Yin, Zhengyang Wang, Tuo Zhao, Jingbo Shang, Chao Zhang, Bing Yin, Xian Li, Jianshu Chen, Shiyang Li
Abstract: Instruction fine-tuning (IFT) elicits instruction following capabilities and steers the behavior of large language models (LLMs) via supervised learning. However, existing models trained on open-source IFT datasets only have the ability to follow instructions from users, and often fail to follow complex role and rules specified by developers, a.k.a. system prompts. The ability to follow these roles and rules is essential for deployment, as it ensures that the model safely interacts with users within developer defined guidelines. To improve such role and rule following ability, we propose \model, an automated data generation pipeline that generates diverse roles and rules from existing IFT instructions, along with corresponding responses. This data can then be used to train models that follow complex system prompts. The models are evaluated on our newly created benchmarks for role and rule following ability, as well as standard instruction-following benchmarks and general NLP tasks. Our framework significantly improves role and rule following capability in LLMs, as evidenced by over 25% increase in pass-rate on rule adherence, i.e. following all requirements, in our experiments with the Alpaca and Ultrachat datasets. Moreover, our models achieves this increase without any regression on popular instruction following benchmarks.
Authors: Michael D. Skarlinski, Sam Cox, Jon M. Laurent, James D. Braza, Michaela Hinks, Michael J. Hammerling, Manvitha Ponnapati, Samuel G. Rodriques, Andrew D. White
Abstract: Language models are known to produce incorrect information, and their accuracy and reliability for scientific research are still in question. We developed a detailed human-AI comparison method to evaluate language models on real-world literature search tasks, including information retrieval, summarization, and contradiction detection. Our findings show that PaperQA2, an advanced language model focused on improving factual accuracy, matches or outperforms subject matter experts on three realistic literature search tasks, with no restrictions on human participants (full internet access, search tools, and time). PaperQA2 generates cited, Wikipedia-style summaries of scientific topics that are significantly more accurate than current human-written Wikipedia entries. We also present LitQA2, a new benchmark for scientific literature research, which shaped the development of PaperQA2 and contributed to its superior performance. Additionally, PaperQA2 identifies contradictions in scientific literature, a challenging task for humans. It finds an average of 2.34 +/- 1.99 contradictions per paper in a random sample of biology papers, with 70% of these contradictions validated by human experts. These results show that language models can now surpass domain experts in important scientific literature tasks.
Authors: Daniel B. Hier, Thanh Son Do, Tayo Obafemi-Ajayi
Abstract: Large language models (LLMs) have shown improved accuracy in phenotype term normalization tasks when augmented with retrievers that suggest candidate normalizations based on term definitions. In this work, we introduce a simplified retriever that enhances LLM accuracy by searching the Human Phenotype Ontology (HPO) for candidate matches using contextual word embeddings from BioBERT without the need for explicit term definitions. Testing this method on terms derived from the clinical synopses of Online Mendelian Inheritance in Man (OMIM), we demonstrate that the normalization accuracy of a state-of-the-art LLM increases from a baseline of 62.3% without augmentation to 90.3% with retriever augmentation. This approach is potentially generalizable to other biomedical term normalization tasks and offers an efficient alternative to more complex retrieval methods.
Authors: Hongyan Chang, Ali Shahin Shamsabadi, Kleomenis Katevas, Hamed Haddadi, Reza Shokri
Abstract: Prior Membership Inference Attacks (MIAs) on pre-trained Large Language Models (LLMs), adapted from classification model attacks, fail due to ignoring the generative process of LLMs across token sequences. In this paper, we present a novel attack that adapts MIA statistical tests to the perplexity dynamics of subsequences within a data point. Our method significantly outperforms prior loss-based approaches, revealing context-dependent memorization patterns in pre-trained LLMs.
Authors: Daniel B. Hier, Thanh Son Do, Tayo Obafemi-Ajayi
Abstract: Term normalization is the process of mapping a term from free text to a standardized concept and its machine-readable code in an ontology. Accurate normalization of terms that capture phenotypic differences between patients and diseases is critical to the success of precision medicine initiatives. A large language model (LLM), such as GPT-4o, can normalize terms to the Human Phenotype Ontology (HPO), but it may retrieve incorrect HPO IDs. Reported accuracy rates for LLMs on these tasks may be inflated due to imbalanced test datasets skewed towards high-frequency terms. In our study, using a comprehensive dataset of 268,776 phenotype annotations for 12,655 diseases from the HPO, GPT-4o achieved an accuracy of 13.1% in normalizing 11,225 unique terms. However, the accuracy was unevenly distributed, with higher-frequency and shorter terms normalized more accurately than lower-frequency and longer terms. Feature importance analysis, using SHAP and permutation methods, identified low-term frequency as the most significant predictor of normalization errors. These findings suggest that training and evaluation datasets for LLM-based term normalization should balance low- and high-frequency terms to improve model performance, particularly for infrequent terms critical to precision medicine.
Authors: Kartikey Doshi, Jimit Shah, Narendra Shekokar
Abstract: We present TheraGen, an advanced AI-powered mental health chatbot utilizing the LLaMA 2 7B model. This approach builds upon recent advancements in language models and transformer architectures. TheraGen provides all-day personalized, compassionate mental health care by leveraging a large dataset of 1 million conversational entries, combining anonymized therapy transcripts, online mental health discussions, and psychological literature, including APA resources. Our implementation employs transfer learning, fine-tuning, and advanced training techniques to optimize performance. TheraGen offers a user-friendly interface for seamless interaction, providing empathetic responses and evidence-based coping strategies. Evaluation results demonstrate high user satisfaction rates, with 94% of users reporting improved mental well-being. The system achieved a BLEU score of 0.67 and a ROUGE score of 0.62, indicating strong response accuracy. With an average response time of 1395 milliseconds, TheraGen ensures real-time, efficient support. While not a replacement for professional therapy, TheraGen serves as a valuable complementary tool, significantly improving user well-being and addressing the accessibility gap in mental health treatments. This paper details TheraGen's architecture, training methodology, ethical considerations, and future directions, contributing to the growing field of AI-assisted mental healthcare and offering a scalable solution to the pressing need for mental health support.
Authors: Neel Rajani, Lilli Kiessling, Aleksandr Ogaltsov, Claus Lang
Abstract: Although powerful, current cutting-edge LLMs may not fulfil the needs of highly specialised sectors. We introduce KodeXv0.1, a family of large language models that outclass GPT-4 in financial question answering. We utilise the base variants of Llama 3.1 8B and 70B and adapt them to the financial domain through a custom training regime. To this end, we collect and process a large number of publicly available financial documents such as earnings calls and business reports. These are used to generate a high-quality, synthetic dataset consisting of Context-Question-Answer triplets which closely mirror real-world financial tasks. Using the train split of this dataset, we perform RAG-aware 4bit LoRA instruction tuning runs of Llama 3.1 base variants to produce KodeX-8Bv0.1 and KodeX-70Bv0.1. We then complete extensive model evaluations using FinanceBench, FinQABench and the withheld test split of our dataset. Our results show that KodeX-8Bv0.1 is more reliable in financial contexts than cutting-edge instruct models in the same parameter regime, surpassing them by up to 9.24%. In addition, it is even capable of outperforming state-of-the-art proprietary models such as GPT-4 by up to 7.07%. KodeX-70Bv0.1 represents a further improvement upon this, exceeding GPT-4's performance on every tested benchmark.
Authors: Baohua Zhang, Yongyi Huang, Wenyao Cui, Huaping Zhang
Abstract: Role-playing is an easy task for Large Language Models (LLMs), as they are skilled at simulating human behaviors. Many current studies have enabled LLMs to generate responses in the tone of a specific role by fine-tuning the models or using specialized prompts. However, it is typically easy to recognize when a role is being played by LLMs. These models tend to perform poorly when confronted with knowledge that the assumed role does not possess, or a question that requires the specific experience or logic of the role to answer. To address this problem and make LLMs act more like real roles, we propose a Thinking Before Speaking (TBS) model in this paper. Unlike other studies, we first extend the data based on the character's real-life scenarios and the historical dialogue, supplementing each pair of dialogue with the character's mindset. Then we add few data points that include elements beyond the role's knowledge, and fine-tune the LLMs. This approach can help LLMs adopt the role's thought process and logic, avoiding responses that fall outside the role's knowledge base. We have also prepared a dataset and evaluation metrics to test these capabilities. Experimental results show that our TBS model can better emulate a role in terms of tone, knowledge, and mindset.
Authors: Asher Sprigler, Alexander Drobek, Keagan Weinstock, Wendpanga Tapsoba, Gavin Childress, Andy Dao, Lucas Gral
Abstract: Large Language Models (LLMs) have increasingly demonstrated the ability to facilitate the development of multi-agent systems that allow the interpretation of thoughts and actions generated by each individual. Promising advancements have also been made in LLM-based interaction with existing worlds, particularly in interacting with simulated environments. This paper aims to integrate both aforementioned topics (agents & world interaction) into a single simulation where multiple agents can work together to solve a problem, modeling how groups of humans can often solve problems better than individuals. By showing whether LLMs demonstrate the synergy of human collaboration, it could lead to advancements in the applications of LLMs. We implemented two simulations: a physical studio apartment with two roommates, and another where agents collaborate to complete a programming task. We provide a multi-agent framework, discuss the performance of the agents in each simulation, and discuss potential future additions.
Authors: Xin Wang, Xinyi Bai
Abstract: Relation extraction as an important natural Language processing (NLP) task is to identify relations between named entities in text. Recently, graph convolutional networks over dependency trees have been widely used to capture syntactic features and achieved attractive performance. However, most existing dependency-based approaches ignore the positive influence of the words outside the dependency trees, sometimes conveying rich and useful information on relation extraction. In this paper, we propose a novel model, Entity-aware Self-attention Contextualized GCN (ESC-GCN), which efficiently incorporates syntactic structure of input sentences and semantic context of sequences. To be specific, relative position self-attention obtains the overall semantic pairwise correlation related to word position, and contextualized graph convolutional networks capture rich intra-sentence dependencies between words by adequately pruning operations. Furthermore, entity-aware attention layer dynamically selects which token is more decisive to make final relation prediction. In this way, our proposed model not only reduces the noisy impact from dependency trees, but also obtains easily-ignored entity-related semantic representation. Extensive experiments on various tasks demonstrate that our model achieves encouraging performance as compared to existing dependency-based and sequence-based models. Specially, our model excels in extracting relations between entities of long sentences.
Authors: Tracy Cai, Wilson Liang, Donte Townes
Abstract: The traditional songwriting process is rather complex and this is evident in the time it takes to produce lyrics that fit the genre and form comprehensive verses. Our project aims to simplify this process with deep learning techniques, thus optimizing the songwriting process and enabling an artist to hit their target audience by staying in genre. Using a dataset of 18,000 songs off Spotify, we developed a unique preprocessing format using tokens to parse lyrics into individual verses. These results were used to train a baseline pretrained seq2seq model, and a LSTM-based neural network models according to song genres. We found that generation yielded higher recall (ROUGE) in the baseline model, but similar precision (BLEU) for both models. Qualitatively, we found that many of the lyrical phrases generated by the original model were still comprehensible and discernible between which genres they fit into, despite not necessarily being the exact the same as the true lyrics. Overall, our results yielded that lyric generation can reasonably be sped up to produce genre-based lyrics and aid in hastening the songwriting process.
Authors: Yihua Cheng, Kuntai Du, Jiayi Yao, Junchen Jiang
Abstract: As the use of large language models (LLMs) expands rapidly, so does the range of knowledge needed to supplement various LLM queries. Thus, enabling flexible and efficient injection of new knowledge in LLM inference is critical. Three high-level options exist: (i) embedding the knowledge in LLM's weights (i.e., fine-tuning), (ii) including the knowledge as a part of LLM's text input (i.e., in-context learning), or (iii) injecting the KV caches of the new knowledge to LLM during prefill. This paper argues that, although fine-tuning and in-context learning are popular, using KV caches as the medium of knowledge could simultaneously enable more modular management of knowledge injection and more efficient LLM serving with low cost and fast response. To realize these benefits, we envision a Knowledge Delivery Network (KDN), a new system component in LLM services that dynamically optimizes the storage, transfer, and composition of KV cache across LLM engines and other compute and storage resources. We believe that, just like content delivery networks (CDNs), such as Akamai, enabled the success of the Internet ecosystem through their efficient data delivery, KDNs will be critical to the success of LLM applications through their efficient knowledge delivery. We have open-sourced a KDN prototype at https://github.com/LMCache/LMCache.
Authors: Christos Fragkathoulas, Odysseas S. Chlapanis
Abstract: This paper introduces a novel task to assess the faithfulness of large language models (LLMs) using local perturbations and self-explanations. Many LLMs often require additional context to answer certain questions correctly. For this purpose, we propose a new efficient alternative explainability technique, inspired by the commonly used leave-one-out approach. Using this approach, we identify the sufficient and necessary parts for the LLM to generate correct answers, serving as explanations. We propose a metric for assessing faithfulness that compares these crucial parts with the self-explanations of the model. Using the Natural Questions dataset, we validate our approach, demonstrating its effectiveness in explaining model decisions and assessing faithfulness.
Authors: Yanick Fratantonio, Luca Invernizzi, Loua Farah, Kurt Thomas, Marina Zhang, Ange Albertini, Francois Galilee, Giancarlo Metitieri, Julien Cretin, Alex Petit-Bianco, David Tao, Elie Bursztein
Abstract: The task of content-type detection -- which entails identifying the data encoded in an arbitrary byte sequence -- is critical for operating systems, development, reverse engineering environments, and a variety of security applications. In this paper, we introduce Magika, a novel AI-powered content-type detection tool. Under the hood, Magika employs a deep learning model that can execute on a single CPU with just 1MB of memory to store the model's weights. We show that Magika achieves an average F1 score of 99% across over a hundred content types and a test set of more than 1M files, outperforming all existing content-type detection tools today. In order to foster adoption and improvements, we open source Magika under an Apache 2 license on GitHub and make our model and training pipeline publicly available. Our tool has already seen adoption by the Gmail email provider for attachment scanning, and it has been integrated with VirusTotal to aid with malware analysis. We note that this paper discusses the first iteration of Magika, and a more recent version already supports more than 200 content types. The interested reader can see the latest development on the Magika GitHub repository, available at https://github.com/google/magika.
Authors: Yi Cui
Abstract: This paper presents a case study of coding tasks by the latest reasoning models of OpenAI, i.e. o1-preview and o1-mini, in comparison with other frontier models. The o1 models deliver SOTA results for WebApp1K, a single-task benchmark. To this end, we introduce WebApp1K-Duo, a harder benchmark doubling number of tasks and test cases. The new benchmark causes the o1 model performances to decline significantly, falling behind Claude 3.5. Moreover, they consistently fail when confronted with atypical yet correct test cases, a trap non-reasoning models occasionally avoid. We hypothesize that the performance variability is due to instruction comprehension. Specifically, the reasoning mechanism boosts performance when all expectations are captured, meanwhile exacerbates errors when key expectations are missed, potentially impacted by input lengths. As such, we argue that the coding success of reasoning models hinges on the top-notch base model and SFT to ensure meticulous adherence to instructions.
Authors: Ioannis Pitsiorlas, George Arvanitakis, Marios Kountouris
Abstract: This work introduces a novel method for enhancing confidence in anomaly detection in Intrusion Detection Systems (IDS) through the use of a Variational Autoencoder (VAE) architecture. By developing a confidence metric derived from latent space representations, we aim to improve the reliability of IDS predictions against cyberattacks. Applied to the NSL-KDD dataset, our approach focuses on binary classification tasks to effectively distinguish between normal and malicious network activities. The methodology demonstrates a significant enhancement in anomaly detection, evidenced by a notable correlation of 0.45 between the reconstruction error and the proposed metric. Our findings highlight the potential of employing VAEs for more accurate and trustworthy anomaly detection in network security.
Authors: Tanya Chutani, Saikiran Bonthu, Pranab Samanta, Nitin Singhal
Abstract: Positron Emission Tomography (PET) /Computed Tomography (CT) is crucial for diagnosing, managing, and planning treatment for various cancers. Developing reliable deep learning models for the segmentation of tumor lesions in PET/CT scans in a multi-tracer multicenter environment, is a critical area of research. Different tracers, such as Fluorodeoxyglucose (FDG) and Prostate-Specific Membrane Antigen (PSMA), have distinct physiological uptake patterns and data from different centers often vary in terms of acquisition protocols, scanner types, and patient populations. Because of this variability, it becomes more difficult to design reliable segmentation algorithms and generalization techniques due to variations in image quality and lesion detectability. To address this challenge, We trained a 3D Residual encoder U-Net within the no new U-Net framework, aiming to generalize the performance of automatic lesion segmentation of whole body PET/CT scans, across different tracers and clinical sites. Further, We explored several preprocessing techniques and ultimately settled on using the Total Segmentator to crop our training data. Additionally, we applied resampling during this process. During inference, we leveraged test-time augmentations and other post-processing techniques to enhance tumor lesion segmentation. Our team currently hold the top position in the Auto-PET III challenge and outperformed the challenge baseline model in the preliminary test set with Dice score of 0.9627.
Authors: Ye Han, Lijun Zhang, Dejian Meng, Xingyu Hu, Songyu Weng
Abstract: To solve the problem of lateral and logitudinal joint decision-making of multi-vehicle cooperative driving for connected and automated vehicles (CAVs), this paper proposes a Monte Carlo tree search (MCTS) method with parallel update for multi-agent Markov game with limited horizon and time discounted setting. By analyzing the parallel actions in the multi-vehicle joint action space in the partial-steady-state traffic flow, the parallel update method can quickly exclude potential dangerous actions, thereby increasing the search depth without sacrificing the search breadth. The proposed method is tested in a large number of randomly generated traffic flow. The experiment results show that the algorithm has good robustness and better performance than the SOTA reinforcement learning algorithms and heuristic methods. The vehicle driving strategy using the proposed algorithm shows rationality beyond human drivers, and has advantages in traffic efficiency and safety in the coordinating zone.
Authors: Yuxuan Hu, Chenwei Zhang, Min Yang, Xiaodan Liang, Chengming Li, Xiping Hu
Abstract: With the rapid development of deep learning methods, there have been many breakthroughs in the field of text classification. Models developed for this task have been shown to achieve high accuracy. However, most of these models are trained using labeled data from seen domains. It is difficult for these models to maintain high accuracy in a new challenging unseen domain, which is directly related to the generalization of the model. In this paper, we study the multi-source Domain Generalization of text classification and propose a framework to use multiple seen domains to train a model that can achieve high accuracy in an unseen domain. Specifically, we propose a multi-source meta-learning Domain Generalization framework to simulate the process of model generalization to an unseen domain, so as to extract sufficient domain-related features. We introduced a memory mechanism to store domain-specific features, which coordinate with the meta-learning framework. Besides, we adopt the novel "jury" mechanism that enables the model to learn sufficient domain-invariant features. Experiments demonstrate that our meta-learning framework can effectively enhance the ability of the model to generalize to an unseen domain and can outperform the state-of-the-art methods on multi-source text classification datasets.
Authors: Yijiang Ma, Tan Chye Cheah
Abstract: This paper implements a new way of solving a problem called the traveling salesman problem (TSP) using quantum genetic algorithm (QGA). We compared how well this new approach works to the traditional method known as a classical genetic algorithm (CGA). The TSP is a well-established challenge in combinatorial optimization where the objective is to find the most efficient path to visit a series of cities, minimizing the total distance, and returning to the starting point. We chose the TSP to test the performance of both algorithms because of its computational complexity and importance in practical applications. We choose the dataset from the international standard library TSPLIB for our experiments. By designing and implementing both algorithms and conducting experiments on various sizes and types of TSP instances, we provide an in-depth analysis of the accuracy of the optimal solution, the number of iterations, the execution time, and the stability of the algorithms for both. The empirical findings indicate that the CGA outperforms the QGA in terms of finding superior solutions more quickly in most of the test instances, especially when the problem size is large. This suggests that although the principle of quantum computing provides a new way to solve complex combinatorial optimisation problems, the implementation of quantum phenomena and the setting of parameters such as the optimal angle for a quantum revolving gate is challenging and need further optimisation to achieve the desired results. Additionally, it is important to note that the QGA has not been tested on real quantum hardware, so its true performance remains unverified. These limitations provide rich opportunities for further research in the future.
Authors: Bangchao Deng, Xin Jing, Tianyue Yang, Bingqing Qu, Philippe Cudre-Mauroux, Dingqi Yang
Abstract: Human trajectory data, which plays a crucial role in various applications such as crowd management and epidemic prevention, is challenging to obtain due to practical constraints and privacy concerns. In this context, synthetic human trajectory data is generated to simulate as close as possible to real-world human trajectories, often under summary statistics and distributional similarities. However, the complexity of human mobility patterns is oversimplified by these similarities (a.k.a. ``Datasaurus''), resulting in intrinsic biases in both generative model design and benchmarks of the generated trajectories. Against this background, we propose MIRAGE, a huMan-Imitative tRAjectory GenErative model designed as a neural Temporal Point Process integrating an Exploration and Preferential Return model. It imitates the human decision-making process in trajectory generation, rather than fitting any specific statistical distributions as traditional methods do, thus avoiding the Datasaurus issue. Moreover, we also propose a comprehensive task-based evaluation protocol beyond Datasaurus to systematically benchmark trajectory generative models on four typical downstream tasks, integrating multiple techniques and evaluation metrics for each task, to comprehensively assess the ultimate utility of the generated trajectories. We conduct a thorough evaluation of MIRAGE on three real-world user trajectory datasets against a sizeable collection of baselines. Results show that compared to the best baselines, MIRAGE-generated trajectory data not only achieves the best statistical and distributional similarities with 59.0-71.5% improvement, but also yields the best performance in the task-based evaluation with 10.9-33.4% improvement.
Authors: Annette Spooner, Mohammad Karimi Moridani, Azadeh Safarchi, Salim Maher, Fatemeh Vafaee, Amany Zekry, Arcot Sowmya
Abstract: The complementary information found in different modalities of patient data can aid in more accurate modelling of a patient's disease state and a better understanding of the underlying biological processes of a disease. However, the analysis of multi-modal, multi-omics data presents many challenges, including high dimensionality and varying size, statistical distribution, scale and signal strength between modalities. In this work we compare the performance of a variety of ensemble machine learning algorithms that are capable of late integration of multi-class data from different modalities. The ensemble methods and their variations tested were i) a voting ensemble, with hard and soft vote, ii) a meta learner, iii) a multi-modal Adaboost model using a hard vote, a soft vote and a meta learner to integrate the modalities on each boosting round, the PB-MVBoost model and a novel application of a mixture of experts model. These were compared to simple concatenation as a baseline. We examine these methods using data from an in-house study on hepatocellular carcinoma (HCC), along with four validation datasets on studies from breast cancer and irritable bowel disease (IBD). Using the area under the receiver operating curve as a measure of performance we develop models that achieve a performance value of up to 0.85 and find that two boosted methods, PB-MVBoost and Adaboost with a soft vote were the overall best performing models. We also examine the stability of features selected, and the size of the clinical signature determined. Finally, we provide recommendations for the integration of multi-modal multi-class data.
Authors: Nilay Kushawaha, Egidio Falotico
Abstract: Continual learning (CL) refers to the ability of an algorithm to continuously and incrementally acquire new knowledge from its environment while retaining previously learned information. A model trained on one data modality often fails when tested with a different modality. A straightforward approach might be to fuse the two modalities by concatenating their features and training the model on the fused data. However, this requires retraining the model from scratch each time it encounters a new domain. In this paper, we introduce a continual learning algorithm capable of incrementally learning different data modalities by leveraging both class-incremental and domain-incremental learning scenarios in an artificial environment where labeled data is scarce, yet non-iid (independent and identical distribution) unlabeled data from the environment is plentiful. The proposed algorithm is efficient and only requires storing prototypes for each class. We evaluate the algorithm's effectiveness on a challenging custom multimodal dataset comprising of tactile data from a soft pneumatic gripper, and visual data from non-stationary images of objects extracted from video sequences. Additionally, we conduct an ablation study on the custom dataset and the Core50 dataset to highlight the contributions of different components of the algorithm. To further demonstrate the robustness of the algorithm, we perform a real-time experiment for object classification using the soft gripper and an external independent camera setup, all synchronized with the Robot Operating System (ROS) framework.
Authors: Jo\~ao Figueiredo, Afonso Carvalho, Daniel Castro, Daniel Gon\c{c}alves, Nuno Santos
Abstract: A vishing attack is a form of social engineering where attackers use phone calls to deceive individuals into disclosing sensitive information, such as personal data, financial information, or security credentials. Attackers exploit the perceived urgency and authenticity of voice communication to manipulate victims, often posing as legitimate entities like banks or tech support. Vishing is a particularly serious threat as it bypasses security controls designed to protect information. In this work, we study the potential for vishing attacks to escalate with the advent of AI. In theory, AI-powered software bots may have the ability to automate these attacks by initiating conversations with potential victims via phone calls and deceiving them into disclosing sensitive information. To validate this thesis, we introduce ViKing, an AI-powered vishing system developed using publicly available AI technology. It relies on a Large Language Model (LLM) as its core cognitive processor to steer conversations with victims, complemented by a pipeline of speech-to-text and text-to-speech modules that facilitate audio-text conversion in phone calls. Through a controlled social experiment involving 240 participants, we discovered that ViKing has successfully persuaded many participants to reveal sensitive information, even those who had been explicitly warned about the risk of vishing campaigns. Interactions with ViKing's bots were generally considered realistic. From these findings, we conclude that tools like ViKing may already be accessible to potential malicious actors, while also serving as an invaluable resource for cyber awareness programs.
Authors: Weijie Zhao, Huajie Shao, Zhaozhuo Xu, Suzhen Duan, Denghui Zhang
Abstract: Exploring the data sources used to train Large Language Models (LLMs) is a crucial direction in investigating potential copyright infringement by these models. While this approach can identify the possible use of copyrighted materials in training data, it does not directly measure infringing risks. Recent research has shifted towards testing whether LLMs can directly output copyrighted content. Addressing this direction, we investigate and assess LLMs' capacity to generate infringing content by providing them with partial information from copyrighted materials, and try to use iterative prompting to get LLMs to generate more infringing content. Specifically, we input a portion of a copyrighted text into LLMs, prompt them to complete it, and then analyze the overlap between the generated content and the original copyrighted material. Our findings demonstrate that LLMs can indeed generate content highly overlapping with copyrighted materials based on these partial inputs.
Authors: Robert Morabito, Sangmitra Madhusudan, Tyler McDonald, Ali Emami
Abstract: Mitigating explicit and implicit biases in Large Language Models (LLMs) has become a critical focus in the field of natural language processing. However, many current methodologies evaluate scenarios in isolation, without considering the broader context or the spectrum of potential biases within each situation. To address this, we introduce the Sensitivity Testing on Offensive Progressions (STOP) dataset, which includes 450 offensive progressions containing 2,700 unique sentences of varying severity that progressively escalate from less to more explicitly offensive. Covering a broad spectrum of 9 demographics and 46 sub-demographics, STOP ensures inclusivity and comprehensive coverage. We evaluate several leading closed- and open-source models, including GPT-4, Mixtral, and Llama 3. Our findings reveal that even the best-performing models detect bias inconsistently, with success rates ranging from 19.3% to 69.8%. We also demonstrate how aligning models with human judgments on STOP can improve model answer rates on sensitive tasks such as BBQ, StereoSet, and CrowS-Pairs by up to 191%, while maintaining or even improving performance. STOP presents a novel framework for assessing the complex nature of biases in LLMs, which will enable more effective bias mitigation strategies and facilitates the creation of fairer language models.
Authors: Julia Watson, Sophia Lee, Barend Beekhuizen, Suzanne Stevenson
Abstract: We study language ideologies in text produced by LLMs through a case study on English gendered language reform (related to role nouns like congressperson/-woman/-man, and singular they). First, we find political bias: when asked to use language that is "correct" or "natural", LLMs use language most similarly to when asked to align with conservative (vs. progressive) values. This shows how LLMs' metalinguistic preferences can implicitly communicate the language ideologies of a particular political group, even in seemingly non-political contexts. Second, we find LLMs exhibit internal inconsistency: LLMs use gender-neutral variants more often when more explicit metalinguistic context is provided. This shows how the language ideologies expressed in text produced by LLMs can vary, which may be unexpected to users. We discuss the broader implications of these findings for value alignment.
Authors: Zhepeng Wang, Runxue Bao, Yawen Wu, Jackson Taylor, Cao Xiao, Feng Zheng, Weiwen Jiang, Shangqian Gao, Yanfu Zhang
Abstract: Pretrained large language models (LLMs) have revolutionized natural language processing (NLP) tasks such as summarization, question answering, and translation. However, LLMs pose significant security risks due to their tendency to memorize training data, leading to potential privacy breaches and copyright infringement. Accurate measurement of this memorization is essential to evaluate and mitigate these potential risks. However, previous attempts to characterize memorization are constrained by either using prefixes only or by prepending a constant soft prompt to the prefixes, which cannot react to changes in input. To address this challenge, we propose a novel method for estimating LLM memorization using dynamic, prefix-dependent soft prompts. Our approach involves training a transformer-based generator to produce soft prompts that adapt to changes in input, thereby enabling more accurate extraction of memorized data. Our method not only addresses the limitations of previous methods but also demonstrates superior performance in diverse experimental settings compared to state-of-the-art techniques. In particular, our method can achieve the maximum relative improvement of 112.75% and 32.26% over the vanilla baseline in terms of discoverable memorization rate for the text generation task and code generation task respectively.
Authors: Slimane Larabi
Abstract: In this paper, we introduce the gated perceptron, an enhancement of the conventional perceptron, which incorporates an additional input computed as the product of the existing inputs. This allows the perceptron to capture non-linear interactions between features, significantly improving its ability to classify and regress on complex datasets. We explore its application in both linear and non-linear regression tasks using the Iris dataset, as well as binary and multi-class classification problems, including the PIMA Indian dataset and Breast Cancer Wisconsin dataset. Our results demonstrate that the gated perceptron can generate more distinct decision regions compared to traditional perceptrons, enhancing its classification capabilities, particularly in handling non-linear data. Performance comparisons show that the gated perceptron competes with state-of-the-art classifiers while maintaining a simple architecture.
Authors: Kunpeng Xu, Lifei Chen, Shengrui Wang
Abstract: Identifying and understanding dynamic concepts in co-evolving sequences is crucial for analyzing complex systems such as IoT applications, financial markets, and online activity logs. These concepts provide valuable insights into the underlying structures and behaviors of sequential data, enabling better decision-making and forecasting. This paper introduces Wormhole, a novel deep representation learning framework that is concept-aware and designed for co-evolving time sequences. Our model presents a self-representation layer and a temporal smoothness constraint to ensure robust identification of dynamic concepts and their transitions. Additionally, concept transitions are detected by identifying abrupt changes in the latent space, signifying a shift to new behavior - akin to passing through a wormhole. This novel mechanism accurately discerns concepts within co-evolving sequences and pinpoints the exact locations of these wormholes, enhancing the interpretability of the learned representations. Experiments demonstrate that this method can effectively segment time series data into meaningful concepts, providing a valuable tool for analyzing complex temporal patterns and advancing the detection of concept drifts.
Authors: Justin Wang, Haimin Hu, Duy Phuong Nguyen, Jaime Fern\'andez Fisac
Abstract: While robust optimal control theory provides a rigorous framework to compute robot control policies that are provably safe, it struggles to scale to high-dimensional problems, leading to increased use of deep learning for tractable synthesis of robot safety. Unfortunately, existing neural safety synthesis methods often lack convergence guarantees and solution interpretability. In this paper, we present Minimax Actors Guided by Implicit Critic Stackelberg (MAGICS), a novel adversarial reinforcement learning (RL) algorithm that guarantees local convergence to a minimax equilibrium solution. We then build on this approach to provide local convergence guarantees for a general deep RL-based robot safety synthesis algorithm. Through both simulation studies on OpenAI Gym environments and hardware experiments with a 36-dimensional quadruped robot, we show that MAGICS can yield robust control policies outperforming the state-of-the-art neural safety synthesis methods.
Authors: Eric Cullhed
Abstract: This article presents an experiment in fine-tuning a pretrained causal language model (Meta's Llama 3.1 8B Instruct) for aiding in three fundamental tasks of philological research: chronological and geographic attribution as well as text restoration in ancient Greek inscriptions and documentary papyri. Using a prompt-based instruct approach, the fine-tuned models surpass the state of the art in key metrics. For inscriptions, the models achieve a lower average character error rate (CER) of 22.5% (vs. 26.3%), while closely matching top-1 accuracy (60.9% vs. 61.8%) and top-20 accuracy (77.5% vs. 78.3%) for sequences up to 10 characters. They also provide a practical advantage by ignoring spaces during reconstruction, aligning better with the scriptio continua typically used in ancient written artifacts. In geographic attribution, the model outperforms previous benchmarks with a top-1 accuracy of 75.0% (vs. 70.8%) and a top-3 accuracy of 83.7% (vs. 82.1%). For dating, it achieves an average deviation of 26.2 years (vs. 29.3) and a median deviation of 1 year (vs. 3) from the actual date range. The models also set new baselines for documentary papyri, with a CER of 16.3%, a top-1 accuracy of 71.3%, and top-20 of 85.0% in text reconstruction; a top-1 accuracy of 66.4% and top-3 of 79.9% in geographic attribution; and, in chronological attribution, a deviation of 21.7 years from the actual termini post/ante quem, with a median deviation of 0 years.
Authors: Vitaliy Kinakh, Slava Voloshynovskiy
Abstract: Generating synthetic tabular data is critical in machine learning, especially when real data is limited or sensitive. Traditional generative models often face challenges due to the unique characteristics of tabular data, such as mixed data types and varied distributions, and require complex preprocessing or large pretrained models. In this paper, we introduce a novel, lossless binary transformation method that converts any tabular data into fixed-size binary representations, and a corresponding new generative model called Binary Diffusion, specifically designed for binary data. Binary Diffusion leverages the simplicity of XOR operations for noise addition and removal and employs binary cross-entropy loss for training. Our approach eliminates the need for extensive preprocessing, complex noise parameter tuning, and pretraining on large datasets. We evaluate our model on several popular tabular benchmark datasets, demonstrating that Binary Diffusion outperforms existing state-of-the-art models on Travel, Adult Income, and Diabetes datasets while being significantly smaller in size.
Authors: Deonna M. Owens, Ryan A. Rossi, Sungchul Kim, Tong Yu, Franck Dernoncourt, Xiang Chen, Ruiyi Zhang, Jiuxiang Gu, Hanieh Deilamsalehy, Nedim Lipka
Abstract: Large Language Models (LLMs) are powerful tools with the potential to benefit society immensely, yet, they have demonstrated biases that perpetuate societal inequalities. Despite significant advancements in bias mitigation techniques using data augmentation, zero-shot prompting, and model fine-tuning, biases continuously persist, including subtle biases that may elude human detection. Recent research has shown a growing interest in multi-LLM approaches, which have been demonstrated to be effective in improving the quality of reasoning and factuality in LLMs. Building on this approach, we propose a novel multi-LLM debiasing framework aimed at reducing bias in LLMs. Our work is the first to introduce and evaluate two distinct approaches within this framework for debiasing LLMs: a centralized method, where the conversation is facilitated by a single central LLM, and a decentralized method, where all models communicate directly. Our findings reveal that our multi-LLM framework significantly reduces bias in LLMs, outperforming the baseline method across several social groups.
Authors: Abhishek Jaiswal, Nisheeth Srivastava
Abstract: Video game playing is an extremely structured domain where algorithmic decision-making can be tested without adverse real-world consequences. While prevailing methods rely on image inputs to avoid the problem of hand-crafting state space representations, this approach systematically diverges from the way humans actually learn to play games. In this paper, we design object-based input representations that generalize well across a number of video games. Using these representations, we evaluate an agent's ability to learn games similar to an infant - with limited world experience, employing simple inductive biases derived from intuitive representations of physics from the real world. Using such biases, we construct an object category representation to be used by a Q-learning algorithm and assess how well it learns to play multiple games based on observed object affordances. Our results suggest that a human-like object interaction setup capably learns to play several video games, and demonstrates superior generalizability, particularly for unfamiliar objects. Further exploring such methods will allow machines to learn in a human-centric way, thus incorporating more human-like learning benefits.
Authors: Samuel Cahyawijaya
Abstract: Natural language processing (NLP) has witnessed a profound impact of large language models (LLMs) that excel in a multitude of tasks. However, the limitation of LLMs in multilingual settings, particularly in underrepresented languages, remains a significant hurdle. This thesis aims to bridge the gap in NLP research and development by focusing on underrepresented languages. A comprehensive evaluation of LLMs is conducted to assess their capabilities in these languages, revealing the challenges of multilingual and multicultural generalization. Addressing the multilingual generalization gap, this thesis proposes data-and-compute-efficient methods to mitigate the disparity in LLM ability in underrepresented languages, allowing better generalization on underrepresented languages without the loss of task generalization ability. The proposed solutions cover cross-lingual continual instruction tuning, retrieval-based cross-lingual in-context learning, and in-context query alignment. Furthermore, a novel method to measure cultural values alignment between LLMs operating in different languages is proposed, ensuring cultural sensitivity and inclusivity. These contributions aim to enhance the multilingual and multicultural alignment of LLMs in underrepresented languages, ultimately advancing the NLP field toward greater equality and inclusiveness.
Authors: Aidan Gilson, Xuguang Ai, Thilaka Arunachalam, Ziyou Chen, Ki Xiong Cheong, Amisha Dave, Cameron Duic, Mercy Kibe, Annette Kaminaka, Minali Prasad, Fares Siddig, Maxwell Singer, Wendy Wong, Qiao Jin, Tiarnan D. L. Keenan, Xia Hu, Emily Y. Chew, Zhiyong Lu, Hua Xu, Ron A. Adelman, Yih-Chung Tham, Qingyu Chen
Abstract: Despite the potential of Large Language Models (LLMs) in medicine, they may generate responses lacking supporting evidence or based on hallucinated evidence. While Retrieval Augment Generation (RAG) is popular to address this issue, few studies implemented and evaluated RAG in downstream domain-specific applications. We developed a RAG pipeline with 70,000 ophthalmology-specific documents that retrieve relevant documents to augment LLMs during inference time. In a case study on long-form consumer health questions, we systematically evaluated the responses including over 500 references of LLMs with and without RAG on 100 questions with 10 healthcare professionals. The evaluation focuses on factuality of evidence, selection and ranking of evidence, attribution of evidence, and answer accuracy and completeness. LLMs without RAG provided 252 references in total. Of which, 45.3% hallucinated, 34.1% consisted of minor errors, and 20.6% were correct. In contrast, LLMs with RAG significantly improved accuracy (54.5% being correct) and reduced error rates (18.8% with minor hallucinations and 26.7% with errors). 62.5% of the top 10 documents retrieved by RAG were selected as the top references in the LLM response, with an average ranking of 4.9. The use of RAG also improved evidence attribution (increasing from 1.85 to 2.49 on a 5-point scale, P<0.001), albeit with slight decreases in accuracy (from 3.52 to 3.23, P=0.03) and completeness (from 3.47 to 3.27, P=0.17). The results demonstrate that LLMs frequently exhibited hallucinated and erroneous evidence in the responses, raising concerns for downstream applications in the medical domain. RAG substantially reduced the proportion of such evidence but encountered challenges.
Authors: Seonghyeon Lee, Suyeon Kim, Joonwon Jang, Heejae Chon, Dongha Lee, Hwanjo Yu
Abstract: We study the code generation behavior of instruction-tuned models built on top of code pre-trained language models when they could access an auxiliary function to implement a function. We design several ways to provide auxiliary functions to the models by adding them to the query or providing a response prefix to incorporate the ability to utilize auxiliary functions with the instruction-following capability. Our experimental results show the effectiveness of combining the base models' auxiliary function utilization ability with the instruction following ability. In particular, the performance of adopting our approaches with the open-sourced language models surpasses that of the recent powerful proprietary language models, i.e., gpt-4o.
Authors: Sarfaroz Yunusov, Hamza Sidat, Ali Emami
Abstract: This study explores the effectiveness of Large Language Models (LLMs) in creating personalized "mirror stories" that reflect and resonate with individual readers' identities, addressing the significant lack of diversity in literature. We present MirrorStories, a corpus of 1,500 personalized short stories generated by integrating elements such as name, gender, age, ethnicity, reader interest, and story moral. We demonstrate that LLMs can effectively incorporate diverse identity elements into narratives, with human evaluators identifying personalized elements in the stories with high accuracy. Through a comprehensive evaluation involving 26 diverse human judges, we compare the effectiveness of MirrorStories against generic narratives. We find that personalized LLM-generated stories not only outscore generic human-written and LLM-generated ones across all metrics of engagement (with average ratings of 4.22 versus 3.37 on a 5-point scale), but also achieve higher textual diversity while preserving the intended moral. We also provide analyses that include bias assessments and a study on the potential for integrating images into personalized stories.
Authors: Kaivalya Rawal, Himabindu Lakkaraju
Abstract: This paper presents a novel technique for incorporating user input when learning and inferring user preferences. When trying to provide users of black-box machine learning models with actionable recourse, we often wish to incorporate their personal preferences about the ease of modifying each individual feature. These recourse finding algorithms usually require an exhaustive set of tuples associating each feature to its cost of modification. Since it is hard to obtain such costs by directly surveying humans, in this paper, we propose the use of the Bradley-Terry model to automatically infer feature-wise costs using non-exhaustive human comparison surveys. We propose that users only provide inputs comparing entire recourses, with all candidate feature modifications, determining which recourses are easier to implement relative to others, without explicit quantification of their costs. We demonstrate the efficient learning of individual feature costs using MAP estimates, and show that these non-exhaustive human surveys, which do not necessarily contain data for each feature pair comparison, are sufficient to learn an exhaustive set of feature costs, where each feature is associated with a modification cost.
Authors: Kevin Li, Fulu Li
Abstract: We use images of cars of a wide range of varieties to compose an image of an animal such as a bird or a lion for the theme of environmental protection to maximize the information about cars in a single composed image and to raise the awareness about environmental challenges. We present a novel way of image interaction with an artistically-composed photomosaic image, in which a simple operation of "click and display" is used to demonstrate the interactive switch between a tile image in a photomosaic image and the corresponding original car image, which will be automatically saved on the Desktop. We build a multimodal custom GPT named TalkMosaic by incorporating car images information and the related knowledge to ChatGPT. By uploading the original car image to TalkMosaic, we can ask questions about the given car image and get the corresponding answers efficiently and effectively such as where to buy the tire in the car image that satisfies high environmental standards. We give an in-depth analysis on how to speed up the inference of multimodal LLM using sparse attention and quantization techniques with presented probabilistic FlashAttention (PrFlashAttention) and Staircase Adaptive Quantization (SAQ) methods. The implemented prototype demonstrates the feasibility and effectiveness of the presented approach.
Authors: Krzysztof Olejniczak, Xingyue Huang, \.Ismail \.Ilkan Ceylan, Mikhail Galkin
Abstract: Traditional query answering over knowledge graphs -- or broadly over relational data -- is one of the most fundamental problems in data management. Motivated by the incompleteness of modern knowledge graphs, a new setup for query answering has emerged, where the goal is to predict answers that do not necessarily appear in the knowledge graph, but are present in its completion. In this work, we propose AnyCQ, a graph neural network model that can classify answers to any conjunctive query on any knowledge graph, following training. At the core of our framework lies a graph neural network model trained using a reinforcement learning objective to answer Boolean queries. Our approach and problem setup differ from existing query answering studies in multiple dimensions. First, we focus on the problem of query answer classification: given a query and a set of possible answers, classify these proposals as true or false relative to the complete knowledge graph. Second, we study the problem of query answer retrieval: given a query, retrieve an answer to the query relative to the complete knowledge graph or decide that no correct solutions exist. Trained on simple, small instances, AnyCQ can generalize to large queries of arbitrary structure, reliably classifying and retrieving answers to samples where existing approaches fail, which is empirically validated on new and challenging benchmarks. Furthermore, we demonstrate that our AnyCQ models effectively transfer to out-of-distribution knowledge graphs, when equipped with a relevant link predictor, highlighting their potential to serve as a general engine for query answering.
Authors: Ehsan Kabir, Jason D. Bakos, David Andrews, Miaoqing Huang
Abstract: Transformer neural networks (TNN) have been widely utilized on a diverse range of applications, including natural language processing (NLP), machine translation, and computer vision (CV). Their widespread adoption has been primarily driven by the exceptional performance of their multi-head self-attention block used to extract key features from sequential data. The multi-head self-attention block is followed by feedforward neural networks, which play a crucial role in introducing non-linearity to assist the model in learning complex patterns. Despite the popularity of TNNs, there has been limited numbers of hardware accelerators targeting these two critical blocks. Most prior works have concentrated on sparse architectures that are not flexible for popular TNN variants. This paper introduces \textit{ProTEA}, a runtime programmable accelerator tailored for the dense computations of most of state-of-the-art transformer encoders. \textit{ProTEA} is designed to reduce latency by maximizing parallelism. We introduce an efficient tiling of large matrices that can distribute memory and computing resources across different hardware components within the FPGA. We provide run time evaluations of \textit{ProTEA} on a Xilinx Alveo U55C high-performance data center accelerator card. Experimental results demonstrate that \textit{ProTEA} can host a wide range of popular transformer networks and achieve near optimal performance with a tile size of 64 in the multi-head self-attention block and 6 in the feedforward networks block when configured with 8 parallel attention heads, 12 layers, and an embedding dimension of 768 on the U55C. Comparative results are provided showing \textit{ProTEA} is 2.5$\times$ faster than an NVIDIA Titan XP GPU. Results also show that it achieves 1.3 -- 2.8$\times$ speed up compared with current state-of-the-art custom designed FPGA accelerators.
Authors: Quanhui Tang, Jingtao Cao
Abstract: Video inpainting enables seamless content removal and replacement within frames, posing ethical and legal risks when misused. To mitigate these risks, detecting manipulated regions in inpainted videos is critical. Previous detection methods often focus solely on the characteristics derived from spatial and temporal dimensions, which limits their effectiveness by overlooking the unique frequency characteristics of different inpainting algorithms. In this paper, we propose the Frequency Domain Insights Network (FDIN), which significantly enhances detection accuracy by incorporating insights from the frequency domain. Our network features an Adaptive Band Selective Response module to discern frequency characteristics specific to various inpainting techniques and a Fast Fourier Convolution-based Attention module for identifying periodic artifacts in inpainted regions. Utilizing 3D ResBlocks for spatiotemporal analysis, FDIN progressively refines detection precision from broad assessments to detailed localization. Experimental evaluations on public datasets demonstrate that FDIN achieves state-of-the-art performance, setting a new benchmark in video inpainting detection.
Authors: Zhiyuan Li, Dongnan Liu, Chaoyi Zhang, Heng Wang, Tengfei Xue, Weidong Cai
Abstract: Recent advancements in Vision-Language (VL) research have sparked new benchmarks for complex visual reasoning, challenging models' advanced reasoning ability. Traditional Vision-Language Models (VLMs) perform well in visual perception tasks while struggling with complex reasoning scenarios. Conversely, Large Language Models (LLMs) demonstrate robust text reasoning capabilities; however, they lack visual acuity. To bridge this gap, we propose Complex Visual Reasoning Large Language Models (CVR-LLM), capitalizing on VLMs' visual perception proficiency and LLMs' extensive reasoning capability. Unlike recent multimodal large language models (MLLMs) that require a projection layer, our approach transforms images into detailed, context-aware descriptions using an iterative self-refinement loop and leverages LLMs' text knowledge for accurate predictions without extra training. We also introduce a novel multi-modal in-context learning (ICL) methodology to enhance LLMs' contextual understanding and reasoning. Additionally, we introduce Chain-of-Comparison (CoC), a step-by-step comparison technique enabling contrasting various aspects of predictions. Our CVR-LLM presents the first comprehensive study across a wide array of complex visual reasoning tasks and achieves SOTA performance among all.
Authors: Yuqing Huang, Rongyang Zhang, Xuesong He, Xuyang Zhi, Hao Wang, Xin Li, Feiyang Xu, Deguang Liu, Huadong Liang, Yi Li, Jian Cui, Zimu Liu, Shijin Wang, Guoping Hu, Guiquan Liu, Qi Liu, Defu Lian, Enhong Chen
Abstract: There is a growing interest in the role that LLMs play in chemistry which lead to an increased focus on the development of LLMs benchmarks tailored to chemical domains to assess the performance of LLMs across a spectrum of chemical tasks varying in type and complexity. However, existing benchmarks in this domain fail to adequately meet the specific requirements of chemical research professionals. To this end, we propose \textbf{\textit{ChemEval}}, which provides a comprehensive assessment of the capabilities of LLMs across a wide range of chemical domain tasks. Specifically, ChemEval identified 4 crucial progressive levels in chemistry, assessing 12 dimensions of LLMs across 42 distinct chemical tasks which are informed by open-source data and the data meticulously crafted by chemical experts, ensuring that the tasks have practical value and can effectively evaluate the capabilities of LLMs. In the experiment, we evaluate 12 mainstream LLMs on ChemEval under zero-shot and few-shot learning contexts, which included carefully selected demonstration examples and carefully designed prompts. The results show that while general LLMs like GPT-4 and Claude-3.5 excel in literature understanding and instruction following, they fall short in tasks demanding advanced chemical knowledge. Conversely, specialized LLMs exhibit enhanced chemical competencies, albeit with reduced literary comprehension. This suggests that LLMs have significant potential for enhancement when tackling sophisticated tasks in the field of chemistry. We believe our work will facilitate the exploration of their potential to drive progress in chemistry. Our benchmark and analysis will be available at {\color{blue} \url{https://github.com/USTC-StarTeam/ChemEval}}.
Authors: Zhenhong Zhang, Jiajing Chen, Weiyan Shi, Lingjie Yi, Chihang Wang, Qian Yu
Abstract: With the rapid development of artificial intelligence technology, especially the increasingly widespread application of question-and-answer systems, high-quality question generation has become a key component in supporting the development of these systems. This article focuses on knowledge-based question generation technology, which aims to enable computers to simulate the human questioning process based on understanding specific texts or knowledge bases. In light of the issues of hallucination and knowledge gaps present in large-scale language models when applied to knowledge-intensive tasks, this paper proposes an enhanced question generation method that incorporates contrastive learning. This method utilizes multiple models to jointly mine domain knowledge and uses contrastive learning to guide the model in reducing noise and hallucinations in generation. Experimental results show that by designing prompts containing contrasting examples, the model's performance in question generation improves considerably, particularly when contrasting instructions and examples are used simultaneously, leading to the highest quality of generated questions and improved accuracy. These results demonstrate that the method proposed in this study, which combines contrasting context and chain-of-thought prompts, can effectively improve both the quality and the practicality of question generation.
Authors: Xiaotong Zhang, Dingcheng Huang, Kamal Youcef-Toumi
Abstract: Human intelligence possesses the ability to effectively focus on important environmental components, which enhances perception, learning, reasoning, and decision-making. Inspired by this cognitive mechanism, we introduced a novel concept termed relevance for Human-Robot Collaboration (HRC). Relevance is defined as the importance of the objects based on the applicability and pertinence of the objects for the human objective or other factors. In this paper, we further developed a novel two-loop framework integrating real-time and asynchronous processing to quantify relevance and apply relevance for safer and more efficient HRC. The asynchronous loop leverages the world knowledge from an LLM and quantifies relevance, and the real-time loop executes scene understanding, human intent prediction, and decision-making based on relevance. In decision making, we proposed and developed a human robot task allocation method based on relevance and a novel motion generation and collision avoidance methodology considering the prediction of human trajectory. Simulations and experiments show that our methodology for relevance quantification can accurately and robustly predict the human objective and relevance, with an average accuracy of up to 0.90 for objective prediction and up to 0.96 for relevance prediction. Moreover, our motion generation methodology reduces collision cases by 63.76% and collision frames by 44.74% when compared with a state-of-the-art (SOTA) collision avoidance method. Our framework and methodologies, with relevance, guide the robot on how to best assist humans and generate safer and more efficient actions for HRC.
Authors: Linxiao Wu, Yuanshuai Luo, Binrong Zhu, Guiran Liu, Rui Wang, Qian Yu
Abstract: Amidst the swift evolution of social media platforms and e-commerce ecosystems, the domain of opinion mining has surged as a pivotal area of exploration within natural language processing. A specialized segment within this field focuses on extracting nuanced evaluations tied to particular elements within textual contexts. This research advances a composite framework that amalgamates the positional cues of topical descriptors. The proposed system converts syntactic structures into a matrix format, leveraging convolutions and attention mechanisms within a graph to distill salient characteristics. Incorporating the positional relevance of descriptors relative to lexical items enhances the sequential integrity of the input. Trials have substantiated that this integrated graph-centric scheme markedly elevates the efficacy of evaluative categorization, showcasing preeminence.
Authors: Ziyan Wang, Bin Liu, Ling Xiang
Abstract: Graph Neural Networks (GNNs) have recently achieved significant success, with a key operation involving the aggregation of information from neighboring nodes. Substantial researchers have focused on defining neighbors for aggregation, predominantly based on observed adjacency matrices. However, in many scenarios, the explicitly given graphs contain noise, which can be amplified during the messages-passing process. Therefore, many researchers have turned their attention to latent graph inference, specifically learning a parametric graph. To mitigate fluctuations in latent graph structure learning, this paper proposes a novel Boolean product-based graph residual connection in GNNs to link the latent graph and the original graph. It computes the Boolean product between the latent graph and the original graph at each layer to correct the learning process. The Boolean product between two adjacency matrices is equivalent to triangle detection. Accordingly, the proposed Boolean product graph neural networks can be interpreted as discovering triangular cliques from the original and the latent graph. We validate the proposed method in benchmark datasets and demonstrate its ability to enhance the performance and robustness of GNNs.
Authors: Panayiotis Christou, Shichu Chen, Xupeng Chen, Parijat Dube
Abstract: Time-series forecasting has seen significant advancements with the introduction of token prediction mechanisms such as multi-head attention. However, these methods often struggle to achieve the same performance as in language modeling, primarily due to the quadratic computational cost and the complexity of capturing long-range dependencies in time-series data. State-space models (SSMs), such as Mamba, have shown promise in addressing these challenges by offering efficient solutions with linear RNNs capable of modeling long sequences with larger context windows. However, there remains room for improvement in accuracy and scalability. We propose the use of Test-Time Training (TTT) modules in a parallel architecture to enhance performance in long-term time series forecasting. Through extensive experiments on standard benchmark datasets, we demonstrate that TTT modules consistently outperform state-of-the-art models, including the Mamba-based TimeMachine, particularly in scenarios involving extended sequence and prediction lengths. Our results show significant improvements in Mean Squared Error (MSE) and Mean Absolute Error (MAE), especially on larger datasets such as Electricity, Traffic, and Weather, underscoring the effectiveness of TTT in capturing long-range dependencies. Additionally, we explore various convolutional architectures within the TTT framework, showing that even simple configurations like 1D convolution with small filters can achieve competitive results. This work sets a new benchmark for time-series forecasting and lays the groundwork for future research in scalable, high-performance forecasting models.
Authors: MohammadReza EskandariNasab, Shah Muhammad Hamdi, Soukaina Filali Boubrahimi
Abstract: Generating time series data using Generative Adversarial Networks (GANs) presents several prevalent challenges, such as slow convergence, information loss in embedding spaces, instability, and performance variability depending on the series length. To tackle these obstacles, we introduce a robust framework aimed at addressing and mitigating these issues effectively. This advanced framework integrates the benefits of an Autoencoder-generated embedding space with the adversarial training dynamics of GANs. This framework benefits from a time series-based loss function and oversight from a supervisory network, both of which capture the stepwise conditional distributions of the data effectively. The generator functions within the latent space, while the discriminator offers essential feedback based on the feature space. Moreover, we introduce an early generation algorithm and an improved neural network architecture to enhance stability and ensure effective generalization across both short and long time series. Through joint training, our framework consistently outperforms existing benchmarks, generating high-quality time series data across a range of real and synthetic datasets with diverse characteristics.
Authors: Sijia Wang, Chen Wang, Zhenhao Zhao, Jiqiang Zhang, Weiran Cai
Abstract: Molecular conformation generation poses a significant challenge in the field of computational chemistry. Recently, Diffusion Probabilistic Models (DPMs) and Score-Based Generative Models (SGMs) are effectively used due to their capacity for generating accurate conformations far beyond conventional physics-based approaches. However, the discrepancy between training and inference rises a critical problem known as the exposure bias. While this issue has been extensively investigated in DPMs, the existence of exposure bias in SGMs and its effective measurement remain unsolved, which hinders the use of compensation methods for SGMs, including ConfGF and Torsional Diffusion as the representatives. In this work, we first propose a method for measuring exposure bias in SGMs used for molecular conformation generation, which confirms the significant existence of exposure bias in these models and measures its value. We design a new compensation algorithm Input Perturbation (IP), which is adapted from a method originally designed for DPMs only. Experimental results show that by introducing IP, SGM-based molecular conformation models can significantly improve both the accuracy and diversity of the generated conformations. Especially by using the IP-enhanced Torsional Diffusion model, we achieve new state-of-the-art performance on the GEOM-Drugs dataset and are on par on GEOM-QM9. We provide the code publicly at https://github.com/jia-975/torsionalDiff-ip.
Authors: MohammadReza EskandariNasab, Shah Muhammad Hamdi, Soukaina Filali Boubrahimi
Abstract: Accurate solar flare prediction is crucial due to the significant risks that intense solar flares pose to astronauts, space equipment, and satellite communication systems. Our research enhances solar flare prediction by utilizing advanced data preprocessing and classification methods on a multivariate time series-based dataset of photospheric magnetic field parameters. First, our study employs a novel preprocessing pipeline that includes missing value imputation, normalization, balanced sampling, near decision boundary sample removal, and feature selection to significantly boost prediction accuracy. Second, we integrate contrastive learning with a GRU regression model to develop a novel classifier, termed ContReg, which employs dual learning methodologies, thereby further enhancing prediction performance. To validate the effectiveness of our preprocessing pipeline, we compare and demonstrate the performance gain of each step, and to demonstrate the efficacy of the ContReg classifier, we compare its performance to that of sequence-based deep learning architectures, machine learning models, and findings from previous studies. Our results illustrate exceptional True Skill Statistic (TSS) scores, surpassing previous methods and highlighting the critical role of precise data preprocessing and classifier development in time series-based solar flare prediction.
Authors: Zhenhua Du, Binbin Xu, Haoyu Zhang, Kai Huo, Shuaifeng Zhi
Abstract: Accurately reconstructing dense and semantically annotated 3D meshes from monocular images remains a challenging task due to the lack of geometry guidance and imperfect view-dependent 2D priors. Though we have witnessed recent advancements in implicit neural scene representations enabling precise 2D rendering simply from multi-view images, there have been few works addressing 3D scene understanding with monocular priors alone. In this paper, we propose MOSE, a neural field semantic reconstruction approach to lift inferred image-level noisy priors to 3D, producing accurate semantics and geometry in both 3D and 2D space. The key motivation for our method is to leverage generic class-agnostic segment masks as guidance to promote local consistency of rendered semantics during training. With the help of semantics, we further apply a smoothness regularization to texture-less regions for better geometric quality, thus achieving mutual benefits of geometry and semantics. Experiments on the ScanNet dataset show that our MOSE outperforms relevant baselines across all metrics on tasks of 3D semantic segmentation, 2D semantic segmentation and 3D surface reconstruction.
Authors: Ling Wang, Chen Wu, Lin Wang
Abstract: Can we directly visualize what we imagine in our brain together with what we describe? The inherent nature of human perception reveals that, when we think, our body can combine language description and build a vivid picture in our brain. Intuitively, generative models should also hold such versatility. In this paper, we introduce BrainDreamer, a novel end-to-end language-guided generative framework that can mimic human reasoning and generate high-quality images from electroencephalogram (EEG) brain signals. Our method is superior in its capacity to eliminate the noise introduced by non-invasive EEG data acquisition and meanwhile achieve a more precise mapping between the EEG and image modality, thus leading to significantly better-generated images. Specifically, BrainDreamer consists of two key learning stages: 1) modality alignment and 2) image generation. In the alignment stage, we propose a novel mask-based triple contrastive learning strategy to effectively align EEG, text, and image embeddings to learn a unified representation. In the generation stage, we inject the EEG embeddings into the pre-trained Stable Diffusion model by designing a learnable EEG adapter to generate high-quality reasoning-coherent images. Moreover, BrainDreamer can accept textual descriptions (e.g., color, position, etc.) to achieve controllable image generation. Extensive experiments show that our method significantly outperforms prior arts in terms of generating quality and quantitative performance.
Authors: Ehsan Kabir, Md. Arafat Kabir, Austin R. J. Downey, Jason D. Bakos, David Andrews, Miaoqing Huang
Abstract: Transformer neural networks (TNNs) are being applied across a widening range of application domains, including natural language processing (NLP), machine translation, and computer vision (CV). Their popularity is largely attributed to the exceptional performance of their multi-head self-attention blocks when analyzing sequential data and extracting features. To date, there are limited hardware accelerators tailored for this mechanism, which is the first step before designing an accelerator for a complete model. This paper proposes \textit{FAMOUS}, a flexible hardware accelerator for dense multi-head attention (MHA) computation of TNNs on field-programmable gate arrays (FPGAs). It is optimized for high utilization of processing elements and on-chip memories to improve parallelism and reduce latency. An efficient tiling of large matrices has been employed to distribute memory and computing resources across different modules on various FPGA platforms. The design is evaluated on Xilinx Alveo U55C and U200 data center cards containing Ultrascale+ FPGAs. Experimental results are presented that show that it can attain a maximum throughput, number of parallel attention heads, embedding dimension and tile size of 328 (giga operations/second (GOPS)), 8, 768 and 64 respectively on the U55C. Furthermore, it is 3.28$\times$ and 2.6$\times$ faster than the Intel Xeon Gold 5220R CPU and NVIDIA V100 GPU respectively. It is also 1.3$\times$ faster than the fastest state-of-the-art FPGA-based accelerator.
Authors: Jason Zhang, Scott Viteri
Abstract: As language models grow more influential and trusted in our society, our ability to reliably steer them toward favorable behaviors becomes increasingly paramount. For this, we investigate the technique of steering vectors: biasing the forward pass of language models using a "steering vector" derived from a specific task. We apply them to steer language models toward performing Chain of Thought (CoT) Reasoning without the need to prompt through natural language. We demonstrate this approach on Llama3 8b and Mistral 7b v0.2, and obtain competitive results compared to CoT-prompted performances on a series of reasoning benchmarks (GSM8k, MMLU, AGI Eval, ARC AI2) and qualitative examples. We find this approach yields consistent steering towards CoT responses and takes less compute than traditional methods of fine-tuning models towards CoT.
Authors: Prasoon Bajpai, Niladri Chatterjee, Subhabrata Dutta, Tanmoy Chakraborty
Abstract: Large Language Models (LLMs) and AI assistants driven by these models are experiencing exponential growth in usage among both expert and amateur users. In this work, we focus on evaluating the reliability of current LLMs as science communicators. Unlike existing benchmarks, our approach emphasizes assessing these models on scientific questionanswering tasks that require a nuanced understanding and awareness of answerability. We introduce a novel dataset, SCiPS-QA, comprising 742 Yes/No queries embedded in complex scientific concepts, along with a benchmarking suite that evaluates LLMs for correctness and consistency across various criteria. We benchmark three proprietary LLMs from the OpenAI GPT family and 13 open-access LLMs from the Meta Llama-2, Llama-3, and Mistral families. While most open-access models significantly underperform compared to GPT-4 Turbo, our experiments identify Llama-3-70B as a strong competitor, often surpassing GPT-4 Turbo in various evaluation aspects. We also find that even the GPT models exhibit a general incompetence in reliably verifying LLM responses. Moreover, we observe an alarming trend where human evaluators are deceived by incorrect responses from GPT-4 Turbo.
Authors: G\"ok\c{c}e Geylan, Jon Paul Janet, Alessandro Tibo, Jiazhen He, Atanas Patronov, Mikhail Kabeshov, Florian David, Werngard Czechtizky, Ola Engkvist, Leonardo De Maria
Abstract: Peptides play a crucial role in the drug design and discovery whether as a therapeutic modality or a delivery agent. Non-natural amino acids (NNAAs) have been used to enhance the peptide properties from binding affinity, plasma stability to permeability. Incorporating novel NNAAs facilitates the design of more effective peptides with improved properties. The generative models used in the field, have focused on navigating the peptide sequence space. The sequence space is formed by combinations of a predefined set of amino acids. However, there is still a need for a tool to explore the peptide landscape beyond this enumerated space to unlock and effectively incorporate de novo design of new amino acids. To thoroughly explore the theoretical chemical space of the peptides, we present PepINVENT, a novel generative AI-based tool as an extension to the small molecule molecular design platform, REINVENT. PepINVENT navigates the vast space of natural and non-natural amino acids to propose valid, novel, and diverse peptide designs. The generative model can serve as a central tool for peptide-related tasks, as it was not trained on peptides with specific properties or topologies. The prior was trained to understand the granularity of peptides and to design amino acids for filling the masked positions within a peptide. PepINVENT coupled with reinforcement learning enables the goal-oriented design of peptides using its chemistry-informed generative capabilities. This study demonstrates PepINVENT's ability to explore the peptide space with unique and novel designs, and its capacity for property optimization in the context of therapeutically relevant peptides. Our tool can be employed for multi-parameter learning objectives, peptidomimetics, lead optimization, and variety of other tasks within the peptide domain.
Authors: Tongxuan Liu, Xingyu Wang, Weizhe Huang, Wenjiang Xu, Yuting Zeng, Lei Jiang, Hailong Yang, Jing Li
Abstract: In recent years, Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse NLP tasks. Extensive research has explored how to enhance the logical reasoning abilities such as Chain-of-Thought, Chain-of-Thought with Self-Consistency, Tree-Of-Thoughts, and multi-agent debates. In the context of multi-agent debates, significant performance improvements can be achieved with an increasing number of agents and debate rounds. However, the escalation in the number of agents and debate rounds can drastically raise the tokens cost of debates, thereby limiting the scalability of the multi-agent debate technique. To better harness the advantages of multi-agent debates in logical reasoning tasks, this paper proposes a method to significantly reduce token cost in multi-agent debates. This approach involves dividing all agents into multiple debate groups, with agents engaging in debates within their respective groups and sharing interim debate results between groups. Comparative experiments across multiple datasets have demonstrated that this method can reduce the total tokens by up to 51.7% during debates and while potentially enhancing accuracy by as much as 25%. Our method significantly enhances the performance and efficiency of interactions in the multi-agent debate.
Authors: Grace Tang, Swetha Rajkumar, Yifei Zhou, Homer Rich Walke, Sergey Levine, Kuan Fang
Abstract: Building generalist robotic systems involves effectively endowing robots with the capabilities to handle novel objects in an open-world setting. Inspired by the advances of large pre-trained models, we propose Keypoint Affordance Learning from Imagined Environments (KALIE), which adapts pre-trained Vision Language Models (VLMs) for robotic control in a scalable manner. Instead of directly producing motor commands, KALIE controls the robot by predicting point-based affordance representations based on natural language instructions and visual observations of the scene. The VLM is trained on 2D images with affordances labeled by humans, bypassing the need for training data collected on robotic systems. Through an affordance-aware data synthesis pipeline, KALIE automatically creates massive high-quality training data based on limited example data manually collected by humans. We demonstrate that KALIE can learn to robustly solve new manipulation tasks with unseen objects given only 50 example data points. Compared to baselines using pre-trained VLMs, our approach consistently achieves superior performance.
Authors: Marcus Kessel, Colin Atkinson
Abstract: Generative AI (GAI) holds great potential to improve software engineering productivity, but its untrustworthy outputs, particularly in code synthesis, pose significant challenges. The need for extensive verification and validation (V&V) of GAI-generated artifacts may undermine the potential productivity gains. This paper proposes a way of mitigating these risks by exploiting GAI's ability to generate multiple versions of code and tests to facilitate comparative analysis across versions. Rather than relying on the quality of a single test or code module, this "differential GAI" (D-GAI) approach promotes more reliable quality evaluation through version diversity. We introduce the Large-Scale Software Observatorium (LASSO), a platform that supports D-GAI by executing and analyzing large sets of code versions and tests. We discuss how LASSO enables rigorous evaluation of GAI-generated artifacts and propose its application in both software development and GAI research.
Authors: Ruilin Luo, Liyuan Wang, Binghuai Lin, Zicheng Lin, Yujiu Yang
Abstract: Large Language Models (LLMs) have emerged as powerful tools for Text-to-SQL tasks, exhibiting remarkable reasoning capabilities. Different from tasks such as math word problems and commonsense reasoning, SQL solutions have a relatively fixed pattern. This facilitates the investigation of whether LLMs can benefit from categorical thinking, mirroring how humans acquire knowledge through inductive reasoning based on comparable examples. In this study, we propose that employing query group partitioning allows LLMs to focus on learning the thought processes specific to a single problem type, consequently enhancing their reasoning abilities across diverse difficulty levels and problem categories. Our experiments reveal that multiple advanced LLMs, when equipped with PTD-SQL, can either surpass or match previous state-of-the-art (SOTA) methods on the Spider and BIRD datasets. Intriguingly, models with varying initial performances have exhibited significant improvements, mainly at the boundary of their capabilities after targeted drilling, suggesting a parallel with human progress. Code is available at https://github.com/lrlbbzl/PTD-SQL.
Authors: Fabio Ferreira, Moreno Schlageter, Raghu Rajan, Andre Biedenkapp, Frank Hutter
Abstract: A World Model is a compressed spatial and temporal representation of a real world environment that allows one to train an agent or execute planning methods. However, world models are typically trained on observations from the real world environment, and they usually do not enable learning policies for other real environments. We propose One-Shot World Model (OSWM), a transformer world model that is learned in an in-context learning fashion from purely synthetic data sampled from a prior distribution. Our prior is composed of multiple randomly initialized neural networks, where each network models the dynamics of each state and reward dimension of a desired target environment. We adopt the supervised learning procedure of Prior-Fitted Networks by masking next-state and reward at random context positions and query OSWM to make probabilistic predictions based on the remaining transition context. During inference time, OSWM is able to quickly adapt to the dynamics of a simple grid world, as well as the CartPole gym and a custom control environment by providing 1k transition steps as context and is then able to successfully train environment-solving agent policies. However, transferring to more complex environments remains a challenge, currently. Despite these limitations, we see this work as an important stepping-stone in the pursuit of learning world models purely from synthetic data.
Authors: Jaehan Kim, Minkyoo Song, Seung Ho Na, Seungwon Shin
Abstract: Parameter-efficient fine-tuning (PEFT) has become a key training strategy for large language models. However, its reliance on fewer trainable parameters poses security risks, such as task-agnostic backdoors. Despite their severe impact on a wide range of tasks, there is no practical defense solution available that effectively counters task-agnostic backdoors within the context of PEFT. In this study, we introduce Obliviate, a PEFT-integrable backdoor defense. We develop two techniques aimed at amplifying benign neurons within PEFT layers and penalizing the influence of trigger tokens. Our evaluations across three major PEFT architectures show that our method can significantly reduce the attack success rate of the state-of-the-art task-agnostic backdoors (83.6%$\downarrow$). Furthermore, our method exhibits robust defense capabilities against both task-specific backdoors and adaptive attacks. Source code will be obtained at https://github.com/obliviateARR/Obliviate.
Authors: Pablo Bernabeu-Perez, Enrique Lopez-Cuena, Dario Garcia-Gasulla
Abstract: The continued release of new and better image generation models increases the demand for synthetic image detectors. In such a dynamic field, detectors need to be able to generalize widely and be robust to uncontrolled alterations. The present work is motivated by this setting, when looking at the role of time, image transformations and data sources, for detector generalization. In these experiments, none of the evaluated detectors is found universal, but results indicate an ensemble could be. Experiments on data collected in the wild show this task to be more challenging than the one defined by large-scale datasets, pointing to a gap between experimentation and actual practice. Finally, we observe a race equilibrium effect, where better generators lead to better detectors, and vice versa. We hypothesize this pushes the field towards a perpetually close race between generators and detectors.
Authors: Yan Zhong, Zhixin Yan, Yi Xie, Shibin Wu, Huaidong Zhang, Lin Shu, Peiru Zhou
Abstract: Diabetic foot neuropathy (DFN) is a critical factor leading to diabetic foot ulcers, which is one of the most common and severe complications of diabetes mellitus (DM) and is associated with high risks of amputation and mortality. Despite its significance, existing datasets do not directly derive from plantar data and lack continuous, long-term foot-specific information. To advance DFN research, we have collected a novel dataset comprising continuous plantar pressure data to recognize diabetic foot neuropathy. This dataset includes data from 94 DM patients with DFN and 41 DM patients without DFN. Moreover, traditional methods divide datasets by individuals, potentially leading to significant domain discrepancies in some feature spaces due to the absence of mid-domain data. In this paper, we propose an effective domain adaptation method to address this proplem. We split the dataset based on convolutional feature statistics and select appropriate sub-source domains to enhance efficiency and avoid negative transfer. We then align the distributions of each source and target domain pair in specific feature spaces to minimize the domain gap. Comprehensive results validate the effectiveness of our method on both the newly proposed dataset for DFN recognition and an existing dataset.
Authors: Blessed Guda, Gabrial Zencha A., Lawrence Francis, Carlee Joe-Wong
Abstract: Large Language models (LLMs) have brought about substantial advancements in the field of Question Answering (QA) systems. These models do remarkably well in addressing intricate inquiries in a variety of disciplines. However, because of domain-specific vocabulary, complex technological concepts, and the requirement for exact responses applying LLMs to specialized sectors like telecommunications presents additional obstacles. GPT-3.5 has been used in recent work, to obtain noteworthy accuracy for telecom-related questions in a Retrieval Augmented Generation (RAG) framework. Notwithstanding these developments, the practical use of models such as GPT-3.5 is restricted by their proprietary nature and high computing demands. This paper introduces QMOS, an innovative approach which uses a Question-Masked loss and Option Shuffling trick to enhance the performance of LLMs in answering Multiple-Choice Questions in the telecommunications domain. Our focus was on using opensource, smaller language models (Phi-2 and Falcon-7B) within an enhanced RAG framework. Our multi-faceted approach involves several enhancements to the whole LLM-RAG pipeline of finetuning, retrieval, prompt engineering and inference. Our approaches significantly outperform existing results, achieving accuracy improvements from baselines of 24.70% to 49.30% with Falcon-7B and from 42.07% to 84.65% with Phi-2.
Authors: Zhihao Lin, Wei Ma, Mingyi Zhou, Yanjie Zhao, Haoyu Wang, Yang Liu, Jun Wang, Li Li
Abstract: In recent years, Large Language Models (LLMs) have gained widespread use, accompanied by increasing concerns over their security. Traditional jailbreak attacks rely on internal model details or have limitations when exploring the unsafe behavior of the victim model, limiting their generalizability. In this paper, we introduce PathSeeker, a novel black-box jailbreak method inspired by the concept of escaping a security maze. This work is inspired by the game of rats escaping a maze. We think that each LLM has its unique "security maze", and attackers attempt to find the exit learning from the received feedback and their accumulated experience to compromise the target LLM's security defences. Our approach leverages multi-agent reinforcement learning, where smaller models collaborate to guide the main LLM in performing mutation operations to achieve the attack objectives. By progressively modifying inputs based on the model's feedback, our system induces richer, harmful responses. During our manual attempts to perform jailbreak attacks, we found that the vocabulary of the response of the target model gradually became richer and eventually produced harmful responses. Based on the observation, we also introduce a reward mechanism that exploits the expansion of vocabulary richness in LLM responses to weaken security constraints. Our method outperforms five state-of-the-art attack techniques when tested across 13 commercial and open-source LLMs, achieving high attack success rates, especially in strongly aligned commercial models like GPT-4o-mini, Claude-3.5, and GLM-4-air with strong safety alignment. This study aims to improve the understanding of LLM security vulnerabilities and we hope that this sturdy can contribute to the development of more robust defenses.
Authors: Rusham Elahi, Zia Tahseen, Tehreem Fatima, Syed Wafa Zahra, Hafiz Muhammad Abubakar, Tehreem Zafar, Aqs Younas, Muhammad Talha Quddoos, Usman Nazir
Abstract: Primary healthcare is a crucial strategy for achieving universal health coverage. South Asian countries are working to improve their primary healthcare system through their country specific policies designed in line with WHO health system framework using the six thematic pillars: Health Financing, Health Service delivery, Human Resource for Health, Health Information Systems, Governance, Essential Medicines and Technology, and an addition area of Cross-Sectoral Linkages. Measuring the current accessibility of healthcare facilities and workforce availability is essential for improving healthcare standards and achieving universal health coverage in developing countries. Data-driven surveillance approaches are required that can provide rapid, reliable, and geographically scalable solutions to understand a) which communities and areas are most at risk of inequitable access and when, b) what barriers to health access exist, and c) how they can be overcome in ways tailored to the specific challenges faced by individual communities. We propose to harness current breakthroughs in Earth-observation (EO) technology, which provide the ability to generate accurate, up-to-date, publicly accessible, and reliable data, which is necessary for equitable access planning and resource allocation to ensure that vaccines, and other interventions reach everyone, particularly those in greatest need, during normal and crisis times. This requires collaboration among countries to identify evidence based solutions to shape health policy and interventions, and drive innovations and research in the region.
Authors: Viet Dung Nguyen, Zhizhuo Yang, Christopher L. Buckley, Alexander Ororbia
Abstract: Although research has produced promising results demonstrating the utility of active inference (AIF) in Markov decision processes (MDPs), there is relatively less work that builds AIF models in the context of environments and problems that take the form of partially observable Markov decision processes (POMDPs). In POMDP scenarios, the agent must infer the unobserved environmental state from raw sensory observations, e.g., pixels in an image. Additionally, less work exists in examining the most difficult form of POMDP-centered control: continuous action space POMDPs under sparse reward signals. In this work, we address issues facing the AIF modeling paradigm by introducing novel prior preference learning techniques and self-revision schedules to help the agent excel in sparse-reward, continuous action, goal-based robotic control POMDP environments. Empirically, we show that our agents offer improved performance over state-of-the-art models in terms of cumulative rewards, relative stability, and success rate. The code in support of this work can be found at https://github.com/NACLab/robust-active-inference.
Authors: Yunfei Ge, Quanyan Zhu
Abstract: Penetration testing is an essential means of proactive defense in the face of escalating cybersecurity incidents. Traditional manual penetration testing methods are time-consuming, resource-intensive, and prone to human errors. Current trends in automated penetration testing are also impractical, facing significant challenges such as the curse of dimensionality, scalability issues, and lack of adaptability to network changes. To address these issues, we propose MEGA-PT, a meta-game penetration testing framework, featuring micro tactic games for node-level local interactions and a macro strategy process for network-wide attack chains. The micro- and macro-level modeling enables distributed, adaptive, collaborative, and fast penetration testing. MEGA-PT offers agile solutions for various security schemes, including optimal local penetration plans, purple teaming solutions, and risk assessment, providing fundamental principles to guide future automated penetration testing. Our experiments demonstrate the effectiveness and agility of our model by providing improved defense strategies and adaptability to changes at both local and network levels.
Authors: Fang Zhang, Shengli Wu
Abstract: The exponential growth of scientific publications in recent years has posed a significant challenge in effective and efficient categorization. This paper introduces a novel approach that combines instance-based learning and ensemble learning techniques for classifying scientific papers into relevant research fields. Working with a classification system with a group of research fields, first a number of typical seed papers are allocated to each of the fields manually. Then for each paper that needs to be classified, we compare it with all the seed papers in every field. Contents and citations are considered separately. An ensemble-based method is then employed to make the final decision. Experimenting with the datasets from DBLP, our experimental results demonstrate that the proposed classification method is effective and efficient in categorizing papers into various research areas. We also find that both content and citation features are useful for the classification of scientific papers.
Authors: Chi Chiu So, Siu Pang Yung
Abstract: Finding solutions to partial differential equations (PDEs) is an important and essential component in many scientific and engineering discoveries. One of the common approaches empowered by deep learning is Physics-informed Neural Networks (PINNs). Recently, a new type of fundamental neural network model, Kolmogorov-Arnold Networks (KANs), has been proposed as a substitute of Multilayer Perceptions (MLPs), and possesses trainable activation functions. To enhance KANs in fitting accuracy, a modification of KANs, so called ReLU-KANs, using "square of ReLU" as the basis of its activation functions has been suggested. In this work, we propose another basis of activation functions, namely, Higher-order-ReLU, which is simpler than the basis of activation functions used in KANs, namely, B-splines; allows efficient KAN matrix operations; and possesses smooth and non-zero higher-order derivatives, essential for physics-informed neural networks. Our detailed experiments on two standard and typical PDEs, namely, the linear Poisson equation and nonlinear Burgers' equation with viscosity, reveal that our proposed Higher-order-ReLU-KANs (HRKANs) achieve the highest fitting accuracy and training robustness and lowest training time significantly among KANs, ReLUKANs and HRKANs.
Authors: Minghai Lu, Benjamin Delaware, Tianyi Zhang
Abstract: Interactive theorem provers such as Coq are powerful tools to formally guarantee the correctness of software. However, using these tools requires significant manual effort and expertise. While Large Language Models (LLMs) have shown promise in automatically generating informal proofs in natural language, they are less effective at generating formal proofs in interactive theorem provers. In this paper, we conduct a formative study to identify common mistakes made by LLMs when asked to generate formal proofs. By analyzing 520 proof generation errors made by GPT-3.5, we found that GPT-3.5 often identified the correct high-level structure of a proof, but struggled to get the lower-level details correct. Based on this insight, we propose PALM, a novel generate-then-repair approach that first prompts an LLM to generate an initial proof and then leverages targeted symbolic methods to iteratively repair low-level problems. We evaluate PALM on a large dataset that includes more than 10K theorems. Our results show that PALM significantly outperforms other state-of-the-art approaches, successfully proving 76.6% to 180.4% more theorems. Moreover, PALM proves 1270 theorems beyond the reach of existing approaches. We also demonstrate the generalizability of PALM across different LLMs.
Authors: Navid Ayoobi, Lily Knab, Wen Cheng, David Pantoja, Hamidreza Alikhani, Sylvain Flamant, Jin Kim, Arjun Mukherjee
Abstract: While large language models (LLMs) exhibit significant utility across various domains, they simultaneously are susceptible to exploitation for unethical purposes, including academic misconduct and dissemination of misinformation. Consequently, AI-generated text detection systems have emerged as a countermeasure. However, these detection mechanisms demonstrate vulnerability to evasion techniques and lack robustness against textual manipulations. This paper introduces back-translation as a novel technique for evading detection, underscoring the need to enhance the robustness of current detection systems. The proposed method involves translating AI-generated text through multiple languages before back-translating to English. We present a model that combines these back-translated texts to produce a manipulated version of the original AI-generated text. Our findings demonstrate that the manipulated text retains the original semantics while significantly reducing the true positive rate (TPR) of existing detection methods. We evaluate this technique on nine AI detectors, including six open-source and three proprietary systems, revealing their susceptibility to back-translation manipulation. In response to the identified shortcomings of existing AI text detectors, we present a countermeasure to improve the robustness against this form of manipulation. Our results indicate that the TPR of the proposed method declines by only 1.85% after back-translation manipulation. Furthermore, we build a large dataset of 720k texts using eight different LLMs. Our dataset contains both human-authored and LLM-generated texts in various domains and writing styles to assess the performance of our method and existing detectors. This dataset is publicly shared for the benefit of the research community.
Authors: Isabele Bittencourt, Aparna S. Varde, Pankaj Lal
Abstract: In this paper, we conduct sentiment analysis on social media data to study mass opinion about offshore wind energy. We adapt three machine learning models, namely, TextBlob, VADER, and SentiWordNet because different functions are provided by each model. TextBlob provides subjectivity analysis as well as polarity classification. VADER offers cumulative sentiment scores. SentiWordNet considers sentiments with reference to context and performs classification accordingly. Techniques in NLP are harnessed to gather meaning from the textual data in social media. Data visualization tools are suitably deployed to display the overall results. This work is much in line with citizen science and smart governance via involvement of mass opinion to guide decision support. It exemplifies the role of Machine Learning and NLP here.
Authors: Yuxuan Zhou, Xien Liu, Chen Ning, Ji Wu
Abstract: In the study, we aim to investigate current LLMs' mastery of medical factual knowledge with a dynamic evaluation schema, which can automatically generate multiple test samples for each medical factual knowledge point. Test samples produced directly by LLMs always introduce factual errors and lack diversity in the manner of knowledge expression. To overcome the drawbacks, here we propose a novel evaluation method, Predicate-text Dual Transformation (PretextTrans), by introducing predicate transformations into the dynamic evaluation schema. Specifically, each medical knowledge point is firstly transformed into a predicate expression; then, the predicate expression derives a series of variants through predicate transformations; lastly, the produced predicate variants are transformed back into textual expressions, resulting in a series of test samples with both factual reliability and expression diversity. Using the proposed PretextTrans method, we systematically investigate 12 well-known LLMs' mastery of medical factual knowledge based on two medical datasets. The comparison results show that current LLMs still have significant deficiencies in fully mastering medical knowledge, which may illustrate why current LLMs still perform unsatisfactorily in real-world medical scenarios despite having achieved considerable performance on public benchmarks. Our proposed method serves as an effective solution for evaluation of LLMs in medical domain and offers valuable insights for developing medical-specific LLMs.
Authors: Xuewen Liu, Zhikai Li, Qingyi Gu
Abstract: Diffusion models have shown excellent performance on various image generation tasks, but the substantial computational costs and huge memory footprint hinder their low-latency applications in real-world scenarios. Quantization is a promising way to compress and accelerate models. Nevertheless, due to the wide range and time-varying activations in diffusion models, existing methods cannot maintain both accuracy and efficiency simultaneously for low-bit quantization. To tackle this issue, we propose DilateQuant, a novel quantization framework for diffusion models that offers comparable accuracy and high efficiency. Specifically, we keenly aware of numerous unsaturated in-channel weights, which can be cleverly exploited to reduce the range of activations without additional computation cost. Based on this insight, we propose Weight Dilation (WD) that maximally dilates the unsaturated in-channel weights to a constrained range through a mathematically equivalent scaling. WD costlessly absorbs the activation quantization errors into weight quantization. The range of activations decreases, which makes activations quantization easy. The range of weights remains constant, which makes model easy to converge in training stage. Considering the temporal network leads to time-varying activations, we design a Temporal Parallel Quantizer (TPQ), which sets time-step quantization parameters and supports parallel quantization for different time steps, significantly improving the performance and reducing time cost. To further enhance performance while preserving efficiency, we introduce a Block-wise Knowledge Distillation (BKD) to align the quantized models with the full-precision models at a block level. The simultaneous training of time-step quantization parameters and weights minimizes the time required, and the shorter backpropagation paths decreases the memory footprint of the quantization process.
Authors: Hung-Ting Su, Ya-Ching Hsu, Xudong Lin, Xiang-Qian Shi, Yulei Niu, Han-Yuan Hsu, Hung-yi Lee, Winston H. Hsu
Abstract: Large language models (LLMs) equipped with chain-of-thoughts (CoT) prompting have shown significant multi-step reasoning capabilities in factual content like mathematics, commonsense, and logic. However, their performance in narrative reasoning, which demands greater abstraction capabilities, remains unexplored. This study utilizes tropes in movie synopses to assess the abstract reasoning abilities of state-of-the-art LLMs and uncovers their low performance. We introduce a trope-wise querying approach to address these challenges and boost the F1 score by 11.8 points. Moreover, while prior studies suggest that CoT enhances multi-step reasoning, this study shows CoT can cause hallucinations in narrative content, reducing GPT-4's performance. We also introduce an Adversarial Injection method to embed trope-related text tokens into movie synopses without explicit tropes, revealing CoT's heightened sensitivity to such injections. Our comprehensive analysis provides insights for future research directions.
Authors: Xu Yan, Yaoting Jiang, Wenyi Liu, Didi Yi, Haoyang Sang, Jianjun Wei
Abstract: This paper explores a new method for time series data analysis, aiming to overcome the limitations of traditional mining techniques when dealing with multidimensional time series data. Time series data are extensively utilized in diverse fields, including backend services for monitoring and optimizing IT infrastructure, medical diagnosis through continuous patient monitoring and health trend analysis, and internet business for tracking user behavior and forecasting sales. However, since the effective information in time series data is often hidden in sequence fragments, the uncertainty of their length, quantity, and morphological variables brings challenges to mining. To this end, this paper proposes a new spatiotemporal feature representation method, which converts multidimensional time series (MTS) into one-dimensional event sequences by transforming spatially varying events, and uses a series of event symbols to represent the spatial structural information of multidimensional coupling in the sequence, which has good interpretability. Then, this paper introduces a variable-length tuple mining method to extract non-redundant key event subsequences in event sequences as spatiotemporal structural features of motion sequences. This method is an unsupervised method that does not rely on large-scale training samples and defines a new model for representing the spatiotemporal structural features of multidimensional time series. The superior performance of the STEM model is verified by pattern classification experiments on a variety of motion sequences. The research results of this paper provide an important theoretical basis and technical support for understanding and predicting human behavior patterns, and have far-reaching practical application value.
Authors: Aysa Xuemo Fan, Arun Balajiee Lekshmi Narayanan, Mohammad Hassany, Jiaze Ke
Abstract: Large Language Models (LLMs) show promise in generating code comments for novice programmers, but their educational effectiveness remains under-evaluated. This study assesses the instructional quality of code comments produced by GPT-4, GPT-3.5-Turbo, and Llama2, compared to expert-developed comments, focusing on their suitability for novices. Analyzing a dataset of ``easy'' level Java solutions from LeetCode, we find that GPT-4 exhibits comparable quality to expert comments in aspects critical for beginners, such as clarity, beginner-friendliness, concept elucidation, and step-by-step guidance. GPT-4 outperforms Llama2 in discussing complexity (chi-square = 11.40, p = 0.001) and is perceived as significantly more supportive for beginners than GPT-3.5 and Llama2 with Mann-Whitney U-statistics = 300.5 and 322.5, p = 0.0017 and 0.0003). This study highlights the potential of LLMs for generating code comments tailored to novice programmers.
Authors: Dominic Schneider, Lutz Rapp
Abstract: Optical fiber amplifiers are key elements in present optical networks. Failures of these components result in high financial loss of income of the network operator as the communication traffic over an affected link is interrupted. Applying Remaining useful lifetime (RUL) prediction in the context of Predictive Maintenance (PdM) to optical fiber amplifiers to predict upcoming system failures at an early stage, so that network outages can be minimized through planning of targeted maintenance actions, ensures reliability and safety. Optical fiber amplifier are complex systems, that work under various operating conditions, which makes correct forecasting a difficult task. Increased monitoring capabilities of systems results in datasets that facilitate the application of data-driven RUL prediction methods. Deep learning models in particular have shown good performance, but generalization based on comparatively small datasets for RUL prediction is difficult. In this paper, we propose Sparse Low-ranked self-Attention Transformer (SLAT) as a novel RUL prediction method. SLAT is based on an encoder-decoder architecture, wherein two parallel working encoders extract features for sensors and time steps. By utilizing the self-attention mechanism, long-term dependencies can be learned from long sequences. The implementation of sparsity in the attention matrix and a low-rank parametrization reduce overfitting and increase generalization. Experimental application to optical fiber amplifiers exemplified on EDFA, as well as a reference dataset from turbofan engines, shows that SLAT outperforms the state-of-the-art methods.
Authors: Peixin Qin, Chen Huang, Yang Deng, Wenqiang Lei, Tat-Seng Chua
Abstract: With the aid of large language models, current conversational recommender system (CRS) has gaining strong abilities to persuade users to accept recommended items. While these CRSs are highly persuasive, they can mislead users by incorporating incredible information in their explanations, ultimately damaging the long-term trust between users and the CRS. To address this, we propose a simple yet effective method, called PC-CRS, to enhance the credibility of CRS's explanations during persuasion. It guides the explanation generation through our proposed credibility-aware persuasive strategies and then gradually refines explanations via post-hoc self-reflection. Experimental results demonstrate the efficacy of PC-CRS in promoting persuasive and credible explanations. Further analysis reveals the reason behind current methods producing incredible explanations and the potential of credible explanations to improve recommendation accuracy.
Authors: Eshagh Kargar, Ville Kyrki
Abstract: Offline reinforcement learning allows training reinforcement learning models on data from live deployments. However, it is limited to choosing the best combination of behaviors present in the training data. In contrast, simulation environments attempting to replicate the live environment can be used instead of the live data, yet this approach is limited by the simulation-to-reality gap, resulting in a bias. In an attempt to get the best of both worlds, we propose a method that combines an imperfect simulation environment with data from the target environment, to train an offline reinforcement learning policy. Our experiments demonstrate that the proposed method outperforms state-of-the-art approaches CQL, MOPO, and COMBO, especially in scenarios with diverse and challenging dynamics, and demonstrates robust behavior across a variety of experimental conditions. The results highlight that using simulator-generated data can effectively enhance offline policy learning despite the sim-to-real gap, when direct interaction with the real-world is not possible.
Authors: Jiachen Zhou, Mingsi Wang, Tianlin Li, Guozhu Meng, Kai Chen
Abstract: Pose-driven human image animation has achieved tremendous progress, enabling the generation of vivid and realistic human videos from just one single photo. However, it conversely exacerbates the risk of image misuse, as attackers may use one available image to create videos involving politics, violence and other illegal content. To counter this threat, we propose Dormant, a novel protection approach tailored to defend against pose-driven human image animation techniques. Dormant applies protective perturbation to one human image, preserving the visual similarity to the original but resulting in poor-quality video generation. The protective perturbation is optimized to induce misextraction of appearance features from the image and create incoherence among the generated video frames. Our extensive evaluation across 8 animation methods and 4 datasets demonstrates the superiority of Dormant over 6 baseline protection methods, leading to misaligned identities, visual distortions, noticeable artifacts, and inconsistent frames in the generated videos. Moreover, Dormant shows effectiveness on 6 real-world commercial services, even with fully black-box access.
Authors: Sven Kruschel, Nico Hambauer, Sven Weinzierl, Sandra Zilker, Mathias Kraus, Patrick Zschech
Abstract: Machine learning is permeating every conceivable domain to promote data-driven decision support. The focus is often on advanced black-box models due to their assumed performance advantages, whereas interpretable models are often associated with inferior predictive qualities. More recently, however, a new generation of generalized additive models (GAMs) has been proposed that offer promising properties for capturing complex, non-linear patterns while remaining fully interpretable. To uncover the merits and limitations of these models, this study examines the predictive performance of seven different GAMs in comparison to seven commonly used machine learning models based on a collection of twenty tabular benchmark datasets. To ensure a fair and robust model comparison, an extensive hyperparameter search combined with cross-validation was performed, resulting in 68,500 model runs. In addition, this study qualitatively examines the visual output of the models to assess their level of interpretability. Based on these results, the paper dispels the misconception that only black-box models can achieve high accuracy by demonstrating that there is no strict trade-off between predictive performance and model interpretability for tabular data. Furthermore, the paper discusses the importance of GAMs as powerful interpretable models for the field of information systems and derives implications for future work from a socio-technical perspective.
Authors: Tzu-Chieh Liu, Chih-Ting Liu, Shao-Yi Chien
Abstract: We propose Pomo3D, a 3D portrait manipulation framework that allows free accessorizing by decomposing and recomposing portraits and accessories. It enables the avatars to attain out-of-distribution (OOD) appearances of simultaneously wearing multiple accessories. Existing methods still struggle to offer such explicit and fine-grained editing; they either fail to generate additional objects on given portraits or cause alterations to portraits (e.g., identity shift) when generating accessories. This restriction presents a noteworthy obstacle as people typically seek to create charming appearances with diverse and fashionable accessories in the virtual universe. Our approach provides an effective solution to this less-addressed issue. We further introduce the Scribble2Accessories module, enabling Pomo3D to create 3D accessories from user-drawn accessory scribble maps. Moreover, we design a bias-conscious mapper to mitigate biased associations present in real-world datasets. In addition to object-level manipulation above, Pomo3D also offers extensive editing options on portraits, including global or local editing of geometry and texture and avatar stylization, elevating 3D editing of neural portraits to a more comprehensive level.
Authors: Raju Gorain, Omkar Salunke
Abstract: The process of landscaping automotive innovation through patent analysis is crucial for Research and Development teams. It aids in comprehending innovation trends, technological advancements, and the latest technologies from competitors. Traditionally, this process required intensive manual efforts. However, with the advent of Large Language Models (LLMs), it can now be automated, leading to faster and more efficient patent categorization & state-of-the-art of inventive concept extraction. This automation can assist various R\&D teams in extracting relevant information from extensive patent databases. This paper introduces a method based on prompt engineering to extract essential information for landscaping. The information includes the problem addressed by the patent, the technology utilized, and the area of innovation within the vehicle ecosystem (such as safety, Advanced Driver Assistance Systems and more).The result demonstrates the implementation of this method to create a landscape of fuel cell technology using open-source patent data. This approach provides a comprehensive overview of the current state of fuel cell technology, offering valuable insights for future research and development in this field.
Authors: Fang Wang (Florence Wong), Hussam Al Hamadi, Ernesto Damiani
Abstract: Malware visualization analysis incorporating with Machine Learning (ML) has been proven to be a promising solution for improving security defenses on different platforms. In this work, we propose an integrated framework for addressing common problems experienced by ML utilizers in developing malware detection systems. Namely, a pictorial presentation system with extensions is designed to preserve the identities of benign/malign samples by encoding each variable into binary digits and mapping them into black and white pixels. A conditional Generative Adversarial Network based model is adopted to produce synthetic images and mitigate issues of imbalance classes. Detection models architected by Convolutional Neural Networks are for validating performances while training on datasets with and without artifactual samples. Result demonstrates accuracy rates of 98.51% and 97.26% for these two training scenarios.
Authors: Pablo Ramirez Amador, Dinarle Milagro Ortega, Arnold Cesarano
Abstract: Pulmonary diseases are a public health problem that requires accurate and fast diagnostic techniques. In this paper, a method based on convolutional neural networks (CNN), Data Augmentation, ResNet50 and Vision Transformers (ViT) is proposed to detect lung pathologies from medical images. A dataset of X-ray images and CT scans of patients with different lung diseases, such as cancer, pneumonia, tuberculosis and fibrosis, is used. The results obtained by the proposed method are compared with those of other existing methods, using performance metrics such as accuracy, sensitivity, specificity and area under the ROC curve. The results show that the proposed method outperforms the other methods in all metrics, achieving an accuracy of 98% and an area under the ROC curve of 99%. It is concluded that the proposed method is an effective and promising tool for the diagnosis of pulmonary pathologies by medical imaging.
Authors: Daoyang Li, Mingyu Jin, Qingcheng Zeng, Haiyan Zhao, Mengnan Du
Abstract: Probing techniques for large language models (LLMs) have primarily focused on English, overlooking the vast majority of the world's languages. In this paper, we extend these probing methods to a multilingual context, investigating the behaviors of LLMs across diverse languages. We conduct experiments on several open-source LLM models, analyzing probing accuracy, trends across layers, and similarities between probing vectors for multiple languages. Our key findings reveal: (1) a consistent performance gap between high-resource and low-resource languages, with high-resource languages achieving significantly higher probing accuracy; (2) divergent layer-wise accuracy trends, where high-resource languages show substantial improvement in deeper layers similar to English; and (3) higher representational similarities among high-resource languages, with low-resource languages demonstrating lower similarities both among themselves and with high-resource languages. These results highlight significant disparities in LLMs' multilingual capabilities and emphasize the need for improved modeling of low-resource languages.
Authors: Sara Monji-Azad, Marvin Kinz, Claudia Scherl, David M\"annle, J\"urgen Hesser, Nikolas L\"ow
Abstract: Non-rigid point cloud registration is a crucial task in computer vision. Evaluating a non-rigid point cloud registration method requires a dataset with challenges such as large deformation levels, noise, outliers, and incompleteness. Despite the existence of several datasets for deformable point cloud registration, the absence of a comprehensive benchmark with all challenges makes it difficult to achieve fair evaluations among different methods. This paper introduces SynBench, a new non-rigid point cloud registration dataset created using SimTool, a toolset for soft body simulation in Flex and Unreal Engine. SynBench provides the ground truth of corresponding points between two point sets and encompasses key registration challenges, including varying levels of deformation, noise, outliers, and incompleteness. To the best of the authors' knowledge, compared to existing datasets, SynBench possesses three particular characteristics: (1) it is the first benchmark that provides various challenges for non-rigid point cloud registration, (2) SynBench encompasses challenges of varying difficulty levels, and (3) it includes ground truth corresponding points both before and after deformation. The authors believe that SynBench enables future non-rigid point cloud registration methods to present a fair comparison of their achievements. SynBench is publicly available at: https://doi.org/10.11588/data/R9IKCF.
Authors: Ruoyu Song, Muslum Ozgur Ozmen, Hyungsub Kim, Antonio Bianchi, Z. Berkay Celik
Abstract: There is a growing interest in integrating Large Language Models (LLMs) with autonomous driving (AD) systems. However, AD systems are vulnerable to attacks against their object detection and tracking (ODT) functions. Unfortunately, our evaluation of four recent LLM agents against ODT attacks shows that the attacks are 63.26% successful in causing them to crash or violate traffic rules due to (1) misleading memory modules that provide past experiences for decision making, (2) limitations of prompts in identifying inconsistencies, and (3) reliance on ground truth perception data. In this paper, we introduce Hudson, a driving reasoning agent that extends prior LLM-based driving systems to enable safer decision making during perception attacks while maintaining effectiveness under benign conditions. Hudson achieves this by first instrumenting the AD software to collect real-time perception results and contextual information from the driving scene. This data is then formalized into a domain-specific language (DSL). To guide the LLM in detecting and making safe control decisions during ODT attacks, Hudson translates the DSL into natural language, along with a list of custom attack detection instructions. Following query execution, Hudson analyzes the LLM's control decision to understand its causal reasoning process. We evaluate the effectiveness of Hudson using a proprietary LLM (GPT-4) and two open-source LLMs (Llama and Gemma) in various adversarial driving scenarios. GPT-4, Llama, and Gemma achieve, on average, an attack detection accuracy of 83. 3%, 63. 6%, and 73. 6%. Consequently, they make safe control decisions in 86.4%, 73.9%, and 80% of the attacks. Our results, following the growing interest in integrating LLMs into AD systems, highlight the strengths of LLMs and their potential to detect and mitigate ODT attacks.
Authors: Chenxu Wang, Ping Jian, Yang Zhen
Abstract: Logical reading comprehension is a challenging task that entails grasping the underlying semantics of text and applying reasoning to deduce the correct answer. Prior researches have primarily focused on enhancing logical reasoning capabilities through Chain-of-Thought (CoT) or data augmentation. However, previous work constructing chain-of-thought rationales concentrates solely on analyzing correct options, neglecting the incorrect alternatives. Addtionally, earlier efforts on data augmentation by altering contexts rely on rule-based methods, which result in generated contexts that lack diversity and coherence. To address these issues, we propose a Premise-Oriented Data Augmentation (PODA) framework. This framework can generate CoT rationales including analyses for both correct and incorrect options, while constructing diverse and high-quality counterfactual contexts from incorrect candidate options. We integrate summarizing premises and identifying premises for each option into rationales. Subsequently, we employ multi-step prompts with identified premises to construct counterfactual context. To facilitate the model's capabilities to better differentiate the reasoning process associated with each option, we introduce a novel thought-path contrastive learning method that compares reasoning paths between the original and counterfactual samples. Experimental results on three representative LLMs demonstrate that our method can improve the baselines substantially across two challenging logical reasoning benchmarks (ReClor and LogiQA 2.0). The data and code are released at https://github.com/lalalamdbf/TPReasoner.
Authors: Gleb Bazhenov, Oleg Platonov, Liudmila Prokhorenkova
Abstract: Tabular machine learning is an important field for industry and science. In this field, table rows are usually treated as independent data samples, but additional information about relations between them is sometimes available and can be used to improve predictive performance. Such information can be naturally modeled with a graph, thus tabular machine learning may benefit from graph machine learning methods. However, graph machine learning models are typically evaluated on datasets with homogeneous node features, which have little in common with heterogeneous mixtures of numerical and categorical features present in tabular datasets. Thus, there is a critical difference between the data used in tabular and graph machine learning studies, which does not allow one to understand how successfully graph models can be transferred to tabular data. To bridge this gap, we propose a new benchmark of diverse graphs with heterogeneous tabular node features and realistic prediction tasks. We use this benchmark to evaluate a vast set of models, including simple methods previously overlooked in the literature. Our experiments show that graph neural networks (GNNs) can indeed often bring gains in predictive performance for tabular data, but standard tabular models also can be adapted to work with graph data by using simple feature preprocessing, which sometimes enables them to compete with and even outperform GNNs. Based on our empirical study, we provide insights for researchers and practitioners in both tabular and graph machine learning fields.
Authors: David Chanin, James Wilken-Smith, Tom\'a\v{s} Dulka, Hardik Bhatnagar, Joseph Bloom
Abstract: Sparse Autoencoders (SAEs) have emerged as a promising approach to decompose the activations of Large Language Models (LLMs) into human-interpretable latents. In this paper, we pose two questions. First, to what extent do SAEs extract monosemantic and interpretable latents? Second, to what extent does varying the sparsity or the size of the SAE affect monosemanticity / interpretability? By investigating these questions in the context of a simple first-letter identification task where we have complete access to ground truth labels for all tokens in the vocabulary, we are able to provide more detail than prior investigations. Critically, we identify a problematic form of feature-splitting we call feature absorption where seemingly monosemantic latents fail to fire in cases where they clearly should. Our investigation suggests that varying SAE size or sparsity is insufficient to solve this issue, and that there are deeper conceptual issues in need of resolution.
Authors: Arjun Raj, Lei Wang, Tom Gedeon
Abstract: Accurately detecting and tracking high-speed, small objects, such as balls in sports videos, is challenging due to factors like motion blur and occlusion. Although recent deep learning frameworks like TrackNetV1, V2, and V3 have advanced tennis ball and shuttlecock tracking, they often struggle in scenarios with partial occlusion or low visibility. This is primarily because these models rely heavily on visual features without explicitly incorporating motion information, which is crucial for precise tracking and trajectory prediction. In this paper, we introduce an enhancement to the TrackNet family by fusing high-level visual features with learnable motion attention maps through a motion-aware fusion mechanism, effectively emphasizing the moving ball's location and improving tracking performance. Our approach leverages frame differencing maps, modulated by a motion prompt layer, to highlight key motion regions over time. Experimental results on the tennis ball and shuttlecock datasets show that our method enhances the tracking performance of both TrackNetV2 and V3. We refer to our lightweight, plug-and-play solution, built on top of the existing TrackNet, as TrackNetV4.
Authors: Zhou Zhang, Dongzeng Tan, Jiaan Wang, Yilong Chen, Jiarong Xu
Abstract: Emojis have gained immense popularity on social platforms, serving as a common means to supplement or replace text. However, existing data mining approaches generally either completely ignore or simply treat emojis as ordinary Unicode characters, which may limit the model's ability to grasp the rich semantic information in emojis and the interaction between emojis and texts. Thus, it is necessary to release the emoji's power in social media data mining. To this end, we first construct a heterogeneous graph consisting of three types of nodes, i.e. post, word and emoji nodes to improve the representation of different elements in posts. The edges are also well-defined to model how these three elements interact with each other. To facilitate the sharing of information among post, word and emoji nodes, we propose a graph pre-train framework for text and emoji co-modeling, which contains two graph pre-training tasks: node-level graph contrastive learning and edge-level link reconstruction learning. Extensive experiments on the Xiaohongshu and Twitter datasets with two types of downstream tasks demonstrate that our approach proves significant improvement over previous strong baseline methods.
Authors: Linxi Wei, Guorui Xiao, Magdalena Balazinska
Abstract: As an important component of data exploration and integration, Column Type Annotation (CTA) aims to label columns of a table with one or more semantic types. With the recent development of Large Language Models (LLMs), researchers have started to explore the possibility of using LLMs for CTA, leveraging their strong zero-shot capabilities. In this paper, we build on this promising work and improve on LLM-based methods for CTA by showing how to use a Knowledge Graph (KG) to augment the context information provided to the LLM. Our approach, called RACOON, combines both pre-trained parametric and non-parametric knowledge during generation to improve LLMs' performance on CTA. Our experiments show that RACOON achieves up to a 0.21 micro F-1 improvement compared against vanilla LLM inference.
Authors: Sheikh Mannan, Paige Hansen, Vivekanand Pandey Vimal, Hannah N. Davies, Paul DiZio, Nikhil Krishnaswamy
Abstract: Spatial disorientation is a leading cause of fatal aircraft accidents. This paper explores the potential of AI agents to aid pilots in maintaining balance and preventing unrecoverable losses of control by offering cues and corrective measures that ameliorate spatial disorientation. A multi-axis rotation system (MARS) was used to gather data from human subjects self-balancing in a spaceflight analog condition. We trained models over this data to create "digital twins" that exemplified performance characteristics of humans with different proficiency levels. We then trained various reinforcement learning and deep learning models to offer corrective cues if loss of control is predicted. Digital twins and assistant models then co-performed a virtual inverted pendulum (VIP) programmed with identical physics. From these simulations, we picked the 5 best-performing assistants based on task metrics such as crash frequency and mean distance from the direction of balance. These were used in a co-performance study with 20 new human subjects performing a version of the VIP task with degraded spatial information. We show that certain AI assistants were able to improve human performance and that reinforcement-learning based assistants were objectively more effective but rated as less trusted and preferable by humans.
Authors: Hongchen Wang, Kangming Li, Scott Ramsay, Yao Fehlis, Edward Kim, Jason Hattrick-Simpers
Abstract: Large Language Models (LLMs) have the potential to revolutionize scientific research, yet their robustness and reliability in domain-specific applications remain insufficiently explored. This study conducts a comprehensive evaluation and robustness analysis of LLMs within the field of materials science, focusing on domain-specific question answering and materials property prediction. Three distinct datasets are used in this study: 1) a set of multiple-choice questions from undergraduate-level materials science courses, 2) a dataset including various steel compositions and yield strengths, and 3) a band gap dataset, containing textual descriptions of material crystal structures and band gap values. The performance of LLMs is assessed using various prompting strategies, including zero-shot chain-of-thought, expert prompting, and few-shot in-context learning. The robustness of these models is tested against various forms of 'noise', ranging from realistic disturbances to intentionally adversarial manipulations, to evaluate their resilience and reliability under real-world conditions. Additionally, the study uncovers unique phenomena of LLMs during predictive tasks, such as mode collapse behavior when the proximity of prompt examples is altered and performance enhancement from train/test mismatch. The findings aim to provide informed skepticism for the broad use of LLMs in materials science and to inspire advancements that enhance their robustness and reliability for practical applications.
Authors: Yiming Zhang, Jianfeng Chi, Hailey Nguyen, Kartikeya Upasani, Daniel M. Bikel, Jason Weston, Eric Michael Smith
Abstract: Text generation has a fundamental limitation almost by definition: there is no taking back tokens that have been generated, even when they are clearly problematic. In the context of language model safety, when a partial unsafe generation is produced, language models by their nature tend to happily keep on generating similarly unsafe additional text. This is in fact how safety alignment of frontier models gets circumvented in the wild, despite great efforts in improving their safety. Deviating from the paradigm of approaching safety alignment as prevention (decreasing the probability of harmful responses), we propose backtracking, a technique that allows language models to "undo" and recover from their own unsafe generation through the introduction of a special [RESET] token. Our method can be incorporated into either SFT or DPO training to optimize helpfulness and harmlessness. We show that models trained to backtrack are consistently safer than baseline models: backtracking Llama-3-8B is four times more safe than the baseline model (6.1\% $\to$ 1.5\%) in our evaluations without regression in helpfulness. Our method additionally provides protection against four adversarial attacks including an adaptive attack, despite not being trained to do so.
Authors: Stefan Haufe, Rick Wilming, Benedict Clark, Rustam Zhumagambetov, Danny Panknin, Ahc\`ene Boubekki
Abstract: The use of machine learning (ML) in critical domains such as medicine poses risks and requires regulation. One requirement is that decisions of ML systems in high-risk applications should be human-understandable. The field of "explainable artificial intelligence" (XAI) seemingly addresses this need. However, in its current form, XAI is unfit to provide quality control for ML; it itself needs scrutiny. Popular XAI methods cannot reliably answer important questions about ML models, their training data, or a given test input. We recapitulate results demonstrating that popular XAI methods systematically attribute importance to input features that are independent of the prediction target. This limits their utility for purposes such as model and data (in)validation, model improvement, and scientific discovery. We argue that the fundamental reason for this limitation is that current XAI methods do not address well-defined problems and are not evaluated against objective criteria of explanation correctness. Researchers should formally define the problems they intend to solve first and then design methods accordingly. This will lead to notions of explanation correctness that can be theoretically verified and objective metrics of explanation performance that can be assessed using ground-truth data.
Authors: Hyunchai Jeong, Adiba Ejaz, Jin Tian, Elias Bareinboim
Abstract: Testing a hypothesized causal model against observational data is a key prerequisite for many causal inference tasks. A natural approach is to test whether the conditional independence relations (CIs) assumed in the model hold in the data. While a model can assume exponentially many CIs (with respect to the number of variables), testing all of them is both impractical and unnecessary. Causal graphs, which encode these CIs in polynomial space, give rise to local Markov properties that enable model testing with a significantly smaller subset of CIs. Model testing based on local properties requires an algorithm to list the relevant CIs. However, existing algorithms for realistic settings with hidden variables and non-parametric distributions can take exponential time to produce even a single CI constraint. In this paper, we introduce the c-component local Markov property (C-LMP) for causal graphs with hidden variables. Since C-LMP can still invoke an exponential number of CIs, we develop a polynomial delay algorithm to list these CIs in poly-time intervals. To our knowledge, this is the first algorithm that enables poly-delay testing of CIs in causal graphs with hidden variables against arbitrary data distributions. Experiments on real-world and synthetic data demonstrate the practicality of our algorithm.
Authors: Tohida Rehman, Debarshi Kumar Sanyal, Samiran Chattopadhyay
Abstract: The title of a research paper communicates in a succinct style the main theme and, sometimes, the findings of the paper. Coming up with the right title is often an arduous task, and therefore, it would be beneficial to authors if title generation can be automated. In this paper, we fine-tune pre-trained and large language models to generate titles of papers from their abstracts. We also use ChatGPT in a zero-shot setting to generate paper titles. The performance of the models is measured with ROUGE, METEOR, MoverScore, BERTScore and SciBERTScore metrics.
Authors: Vieloszynski Alexis, Soumaya Cherkaoui, Jean-Fr\'ed\'eric Laprade, Oliver Nahman-L\'evesque, Abdallah Aaraba, Shengrui Wang
Abstract: Quantum machine learning consists in taking advantage of quantum computations to generate classical data. A potential application of quantum machine learning is to harness the power of quantum computers for generating classical data, a process essential to a multitude of applications such as enriching training datasets, anomaly detection, and risk management in finance. Given the success of Generative Adversarial Networks in classical image generation, the development of its quantum versions has been actively conducted. However, existing implementations on quantum computers often face significant challenges, such as scalability and training convergence issues. To address these issues, we propose LatentQGAN, a novel quantum model that uses a hybrid quantum-classical GAN coupled with an autoencoder. Although it was initially designed for image generation, the LatentQGAN approach holds potential for broader application across various practical data generation tasks. Experimental outcomes on both classical simulators and noisy intermediate scale quantum computers have demonstrated significant performance enhancements over existing quantum methods, alongside a significant reduction in quantum resources overhead.
Authors: Sangwon Kim, Dasom Ahn, Byoung Chul Ko, In-su Jang, Kwang-Ju Kim
Abstract: The demand for reliable AI systems has intensified the need for interpretable deep neural networks. Concept bottleneck models (CBMs) have gained attention as an effective approach by leveraging human-understandable concepts to enhance interpretability. However, existing CBMs face challenges due to deterministic concept encoding and reliance on inconsistent concepts, leading to inaccuracies. We propose EQ-CBM, a novel framework that enhances CBMs through probabilistic concept encoding using energy-based models (EBMs) with quantized concept activation vectors (qCAVs). EQ-CBM effectively captures uncertainties, thereby improving prediction reliability and accuracy. By employing qCAVs, our method selects homogeneous vectors during concept encoding, enabling more decisive task performance and facilitating higher levels of human intervention. Empirical results using benchmark datasets demonstrate that our approach outperforms the state-of-the-art in both concept and task accuracy.
Authors: Amin Ghafourian, Zhongying CuiZhu, Debo Shi, Ian Chuang, Francois Charette, Rithik Sachdeva, Iman Soltani
Abstract: Human navigation is facilitated through the association of actions with landmarks, tapping into our ability to recognize salient features in our environment. Consequently, navigational instructions for humans can be extremely concise, such as short verbal descriptions, indicating a small memory requirement and no reliance on complex and overly accurate navigation tools. Conversely, current autonomous navigation schemes rely on accurate positioning devices and algorithms as well as extensive streams of sensory data collected from the environment. Inspired by this human capability and motivated by the associated technological gap, in this work we propose a hierarchical end-to-end meta-learning scheme that enables a mobile robot to navigate in a previously unknown environment upon presentation of only a few sample images of a set of landmarks along with their corresponding high-level navigation actions. This dramatically simplifies the wayfinding process and enables easy adoption to new environments. For few-shot waypoint detection, we implement a metric-based few-shot learning technique through distribution embedding. Waypoint detection triggers the multi-task low-level maneuver controller module to execute the corresponding high-level navigation action. We demonstrate the effectiveness of the scheme using a small-scale autonomous vehicle on novel indoor navigation tasks in several previously unseen environments.
Authors: Marissa Radensky, Simra Shahid, Raymond Fok, Pao Siangliulue, Tom Hope, Daniel S. Weld
Abstract: The scientific ideation process often involves blending salient aspects of existing papers to create new ideas. To see if large language models (LLMs) can assist this process, we contribute Scideator, a novel mixed-initiative tool for scientific ideation. Starting from a user-provided set of papers, Scideator extracts key facets (purposes, mechanisms, and evaluations) from these and relevant papers, allowing users to explore the idea space by interactively recombining facets to synthesize inventive ideas. Scideator also helps users to gauge idea novelty by searching the literature for potential overlaps and showing automated novelty assessments and explanations. To support these tasks, Scideator introduces four LLM-powered retrieval-augmented generation (RAG) modules: Analogous Paper Facet Finder, Faceted Idea Generator, Idea Novelty Checker, and Idea Novelty Iterator. In a within-subjects user study, 19 computer-science researchers identified significantly more interesting ideas using Scideator compared to a strong baseline combining a scientific search engine with LLM interaction.
Authors: Humza Wajid Hameed, Geraldin Nanfack, Eugene Belilovsky
Abstract: Spurious correlations are a major source of errors for machine learning models, in particular when aiming for group-level fairness. It has been recently shown that a powerful approach to combat spurious correlations is to re-train the last layer on a balanced validation dataset, isolating robust features for the predictor. However, key attributes can sometimes be discarded by neural networks towards the last layer. In this work, we thus consider retraining a classifier on a set of features derived from all layers. We utilize a recently proposed feature selection strategy to select unbiased features from all the layers. We observe this approach gives significant improvements in worst-group accuracy on several standard benchmarks.
Authors: Zixiang Xian, Chenhui Cui, Rubing Huang, Chunrong Fang, Zhenyu Chen
Abstract: Regarding software engineering (SE) tasks, Large language models (LLMs) have the capability of zero-shot learning, which does not require training or fine-tuning, unlike pre-trained models (PTMs). However, LLMs are primarily designed for natural language output, and cannot directly produce intermediate embeddings from source code. They also face some challenges, for example, the restricted context length may prevent them from handling larger inputs, limiting their applicability to many SE tasks; while hallucinations may occur when LLMs are applied to complex downstream tasks. Motivated by the above facts, we propose zsLLMCode, a novel approach that generates functional code embeddings using LLMs. Our approach utilizes LLMs to convert source code into concise summaries through zero-shot learning, which is then transformed into functional code embeddings using specialized embedding models. This unsupervised approach eliminates the need for training and addresses the issue of hallucinations encountered with LLMs. To the best of our knowledge, this is the first approach that combines LLMs and embedding models to generate code embeddings. We conducted experiments to evaluate the performance of our approach. The results demonstrate the effectiveness and superiority of our approach over state-of-the-art unsupervised methods.
Authors: Taihang Wang, Xiaoman Xu, Yimin Wang, Ye Jiang
Abstract: Real-world applications of large language models (LLMs) in computational social science (CSS) tasks primarily depend on the effectiveness of instruction tuning (IT) or in-context learning (ICL). While IT has shown highly effective at fine-tuning LLMs for various tasks, ICL offers a rapid alternative for task adaptation by learning from examples without explicit gradient updates. In this paper, we evaluate the classification performance of LLMs using IT versus ICL in few-shot CSS tasks. The experimental results indicate that ICL consistently outperforms IT in most CSS tasks. Additionally, we investigate the relationship between the increasing number of training samples and LLM performance. Our findings show that simply increasing the number of samples without considering their quality does not consistently enhance the performance of LLMs with either ICL or IT and can sometimes even result in a performance decline. Finally, we compare three prompting strategies, demonstrating that ICL is more effective than zero-shot and Chain-of-Thought (CoT). Our research highlights the significant advantages of ICL in handling CSS tasks in few-shot settings and emphasizes the importance of optimizing sample quality and prompting strategies to improve LLM classification performance. The code will be made available.
Authors: Hojun Son, Arpan Kusari
Abstract: Domain adaptation for object detection (DAOD) aims to transfer a trained model from a source to a target domain. Various DAOD methods exist, some of which minimize context bias between foreground-background associations in various domains. However, no prior work has studied context bias in DAOD by analyzing changes in background features during adaptation and how context bias is represented in different domains. Our research experiment highlights the potential usability of context bias in DAOD. We address the problem by varying activation values over different layers of trained models and by masking the background, both of which impact the number and quality of detections. We then use one synthetic dataset from CARLA and two different versions of real open-source data, Cityscapes and Cityscapes foggy, as separate domains to represent and quantify context bias. We utilize different metrics such as Maximum Mean Discrepancy (MMD) and Maximum Variance Discrepancy (MVD) to find the layer-specific conditional probability estimates of foreground given manipulated background regions for separate domains. We demonstrate through detailed analysis that understanding of the context bias can affect DAOD approach and foc
Authors: Benjamin Clavi\'e, Antoine Chaffin, Griffin Adams
Abstract: Over the last few years, multi-vector retrieval methods, spearheaded by ColBERT, have become an increasingly popular approach to Neural IR. By storing representations at the token level rather than at the document level, these methods have demonstrated very strong retrieval performance, especially in out-of-domain settings. However, the storage and memory requirements necessary to store the large number of associated vectors remain an important drawback, hindering practical adoption. In this paper, we introduce a simple clustering-based token pooling approach to aggressively reduce the number of vectors that need to be stored. This method can reduce the space & memory footprint of ColBERT indexes by 50% with virtually no retrieval performance degradation. This method also allows for further reductions, reducing the vector count by 66%-to-75% , with degradation remaining below 5% on a vast majority of datasets. Importantly, this approach requires no architectural change nor query-time processing, and can be used as a simple drop-in during indexation with any ColBERT-like model.
Authors: Jingtao Cao, Zheng Zhang, Hongru Wang, Kam-Fai Wong
Abstract: Progress in Text-to-Image (T2I) models has significantly improved the generation of images from textual descriptions. However, existing evaluation metrics do not adequately assess the models' ability to handle a diverse range of textual prompts, which is crucial for their generalizability. To address this, we introduce a new metric called Visual Language Evaluation Understudy (VLEU). VLEU uses large language models to sample from the visual text domain, the set of all possible input texts for T2I models, to generate a wide variety of prompts. The images generated from these prompts are evaluated based on their alignment with the input text using the CLIP model.VLEU quantifies a model's generalizability by computing the Kullback-Leibler divergence between the marginal distribution of the visual text and the conditional distribution of the images generated by the model. This metric provides a quantitative way to compare different T2I models and track improvements during model finetuning. Our experiments demonstrate the effectiveness of VLEU in evaluating the generalization capability of various T2I models, positioning it as an essential metric for future research in text-to-image synthesis.
Authors: Ernie Chang, Pin-Jie Lin, Yang Li, Changsheng Zhao, Daeil Kim, Rastislav Rabatin, Zechun Liu, Yangyang Shi, Vikas Chandra
Abstract: Language model pretraining generally targets a broad range of use cases and incorporates data from diverse sources. However, there are instances where we desire a model that excels in specific areas without markedly compromising performance in other areas. A cost-effective and straightforward approach is sampling with low-dimensional data features, which allows to select large-scale pretraining data for domain-specific use cases. In this work, we revisit importance sampling with n-gram features consisting of multi-granular tokens, which strikes a good balance between sentence compression and representation capabilities. We observed the sampled data to have a high correlation with the target downstream task performance while preserving its effectiveness on other tasks. This leads to the proposed data sampling paradigm where language models can be pretrained more efficiently on selected documents. On eight benchmarks we demonstrate with $\sim$1% of the data, pretrained models perform on par with the full RefinedWeb data and outperform randomly selected samples for model sizes ranging from 125M to 1.5B.
Authors: Yihong Tang, Jiao Ou, Che Liu, Fuzheng Zhang, Di Zhang, Kun Gai
Abstract: Role-playing is an emerging application in the field of Human-Computer Interaction (HCI), primarily implemented through the alignment training of a large language model (LLM) with assigned characters. Despite significant progress, role-playing agents (RPLAs) still struggle with maintaining role-consistency across conversations, particularly when confronted with boundary queries subtly related to character attributes. In this paper, we present ERABAL, a framework aimed at enhancing RPLAs' role-playing capabilities through boundary-aware learning. ERABAL encompasses a generation pipeline for role-specific dialogues and a concomitant methodology for alignment training. Through comprehensive evaluations, we demonstrate that ERABAL is both efficient and effective. By training with significantly fewer dialogues than those used in leading approaches, ERABAL achieves notable improvements across WikiRoleEval, CharacterEval, and the role-playing subset of MT-Bench compared to the generalist baseline models. Our code and datasets will be made publicly available to support further research.
Authors: Ashish Prasad, Pranav Jeevan, Amit Sethi
Abstract: Current video summarization methods largely rely on transformer-based architectures, which, due to their quadratic complexity, require substantial computational resources. In this work, we address these inefficiencies by enhancing the Direct-to-Summarize Network (DSNet) with more resource-efficient token mixing mechanisms. We show that replacing traditional attention with alternatives like Fourier, Wavelet transforms, and Nystr\"omformer improves efficiency and performance. Furthermore, we explore various pooling strategies within the Regional Proposal Network, including ROI pooling, Fast Fourier Transform pooling, and flat pooling. Our experimental results on TVSum and SumMe datasets demonstrate that these modifications significantly reduce computational costs while maintaining competitive summarization performance. Thus, our work offers a more scalable solution for video summarization tasks.
Authors: Jiahao Yu, Yangguang Shao, Hanwen Miao, Junzheng Shi, Xinyu Xing
Abstract: Large Language Models (LLMs) have gained widespread use in various applications due to their powerful capability to generate human-like text. However, prompt injection attacks, which involve overwriting a model's original instructions with malicious prompts to manipulate the generated text, have raised significant concerns about the security and reliability of LLMs. Ensuring that LLMs are robust against such attacks is crucial for their deployment in real-world applications, particularly in critical tasks. In this paper, we propose PROMPTFUZZ, a novel testing framework that leverages fuzzing techniques to systematically assess the robustness of LLMs against prompt injection attacks. Inspired by software fuzzing, PROMPTFUZZ selects promising seed prompts and generates a diverse set of prompt injections to evaluate the target LLM's resilience. PROMPTFUZZ operates in two stages: the prepare phase, which involves selecting promising initial seeds and collecting few-shot examples, and the focus phase, which uses the collected examples to generate diverse, high-quality prompt injections. Using PROMPTFUZZ, we can uncover more vulnerabilities in LLMs, even those with strong defense prompts. By deploying the generated attack prompts from PROMPTFUZZ in a real-world competition, we achieved the 7th ranking out of over 4000 participants (top 0.14%) within 2 hours. Additionally, we construct a dataset to fine-tune LLMs for enhanced robustness against prompt injection attacks. While the fine-tuned model shows improved robustness, PROMPTFUZZ continues to identify vulnerabilities, highlighting the importance of robust testing for LLMs. Our work emphasizes the critical need for effective testing tools and provides a practical framework for evaluating and improving the robustness of LLMs against prompt injection attacks.
Authors: Zheng Hui, Zhaoxiao Guo, Hang Zhao, Juanyong Duan, Congrui Huang
Abstract: In different NLP tasks, detecting harmful content is crucial for online environments, especially with the growing influence of social media. However, previous research has two main issues: 1) a lack of data in low-resource settings, and 2) inconsistent definitions and criteria for judging harmful content, requiring classification models to be robust to spurious features and diverse. We propose Toxicraft, a novel framework for synthesizing datasets of harmful information to address these weaknesses. With only a small amount of seed data, our framework can generate a wide variety of synthetic, yet remarkably realistic, examples of toxic information. Experimentation across various datasets showcases a notable enhancement in detection model robustness and adaptability, surpassing or close to the gold labels. We release the generated data at Github upon acceptance.
Authors: Jianqi Zhang, Mengxuan Wang, Jingyao Wang, Lingyu Si, Changwen Zheng, Fanjiang Xu
Abstract: Scene recognition, particularly for aerial and underwater images, often suffers from various types of degradation, such as blurring or overexposure. Previous works that focus on convolutional neural networks have been shown to be able to extract panoramic semantic features and perform well on scene recognition tasks. However, low-quality images still impede model performance due to the inappropriate use of high-level semantic features. To address these To address these challenges, we propose an adaptive selection mechanism to identify the most important and robust regions with high-level features. Thus, the model can perform learning via these regions to avoid interference. implement a learnable mask in the neural network, which can filter high-level features by assigning weights to different regions of the feature matrix. We also introduce a regularization term to further enhance the significance of key high-level feature regions. Different from previous methods, our learnable matrix pays extra attention to regions that are important to multiple categories but may cause misclassification and sets constraints to reduce the influence of such regions.This is a plug-and-play architecture that can be easily extended to other methods. Additionally, we construct an Underwater Geological Scene Classification dataset to assess the effectiveness of our model. Extensive experimental results demonstrate the superiority and robustness of our proposed method over state-of-the-art techniques on two datasets.
Authors: Dasol Choi, Dongbin Na
Abstract: With the explosive growth of deep learning applications, the right to be forgotten has become increasingly in demand in various AI industries. For example, given a facial recognition system, some individuals may wish to remove images that might have been used in the training phase from the trained model. Unfortunately, modern deep neural networks sometimes unexpectedly leak personal identities. Recent studies have presented various machine unlearning algorithms to make a trained model unlearn the data to be forgotten. While these methods generally perform well in terms of forgetting scores, we have found that an unexpected modelutility drop can occur. This phenomenon, which we term correlation collapse, happens when the machine unlearning algorithms reduce the useful correlation between image features and the true label. To address this challenge, we propose Distribution-Level Feature Distancing (DLFD), a novel method that efficiently forgets instances while preventing correlation collapse. Our method synthesizes data samples so that the generated data distribution is far from the distribution of samples being forgotten in the feature space, achieving effective results within a single training epoch. Through extensive experiments on facial recognition datasets, we demonstrate that our approach significantly outperforms state-of-the-art machine unlearning methods.
Authors: Haocheng Zhao, Runwei Guan, Taoyu Wu, Ka Lok Man, Limin Yu, Yutao Yue
Abstract: 4D millimeter-wave (MMW) radar, which provides both height information and dense point cloud data over 3D MMW radar, has become increasingly popular in 3D object detection. In recent years, radar-vision fusion models have demonstrated performance close to that of LiDAR-based models, offering advantages in terms of lower hardware costs and better resilience in extreme conditions. However, many radar-vision fusion models treat radar as a sparse LiDAR, underutilizing radar-specific information. Additionally, these multi-modal networks are often sensitive to the failure of a single modality, particularly vision. To address these challenges, we propose the Radar Depth Lift-Splat-Shoot (RDL) module, which integrates radar-specific data into the depth prediction process, enhancing the quality of visual Bird-Eye View (BEV) features. We further introduce a Unified Feature Fusion (UFF) approach that extracts BEV features across different modalities using shared module. To assess the robustness of multi-modal models, we develop a novel Failure Test (FT) ablation experiment, which simulates vision modality failure by injecting Gaussian noise. We conduct extensive experiments on the View-of-Delft (VoD) and TJ4D datasets. The results demonstrate that our proposed Unified BEVFusion (UniBEVFusion) network significantly outperforms state-of-the-art models on the TJ4D dataset, with improvements of 1.44 in 3D and 1.72 in BEV object detection accuracy.
Authors: Nam Hyeon-Woo, Moon Ye-Bin, Wonseok Choi, Lee Hyun, Tae-Hyun Oh
Abstract: Vision language models (VLMs) have shown promising reasoning capabilities across various benchmarks; however, our understanding of their visual perception remains limited. In this work, we propose an eye examination process to investigate how a VLM perceives images, specifically focusing on key elements of visual recognition, from primitive color and shape to semantic levels. To this end, we introduce a dataset named LENS to guide a VLM to follow the examination and check its readiness. Once the model is ready, we conduct the examination. Through this examination, we quantify and visualize VLMs' sensitivities to color and shape, and semantic matching. Our findings reveal that VLMs have varying sensitivity to different colors while consistently showing insensitivity to green across different VLMs. Also, we found different shape sensitivity and semantic recognition depending on LLM's capacity despite using the same fixed visual encoder. Our analyses and findings have potential to inspire the design of VLMs and the pre-processing of visual input to VLMs for improving application performance.
Authors: Yuyan Chen, Tianhao Yu, Yueze Li, Songzhou Yan, Sijia Liu, Jiaqing Liang, Yanghua Xiao
Abstract: The evaluation of the problem-solving capability under incomplete information scenarios of Large Language Models (LLMs) is increasingly important, encompassing capabilities such as questioning, knowledge search, error detection, and path planning. Current research mainly focus on LLMs' problem-solving capability such as ``Twenty Questions''. However, these kinds of games do not require recognizing misleading cues which are necessary in the incomplete information scenario. Moreover, the existing game such as ``Who is undercover'' are highly subjective, making it challenging for evaluation. Therefore, in this paper, we introduce a novel game named BrainKing based on the ``Who is undercover'' and ``Twenty Questions'' for evaluating LLM capabilities under incomplete information scenarios. It requires LLMs to identify target entities with limited yes-or-no questions and potential misleading answers. By setting up easy, medium, and hard difficulty modes, we comprehensively assess the performance of LLMs across various aspects. Our results reveal the capabilities and limitations of LLMs in BrainKing, providing significant insights of LLM problem-solving levels.
Authors: Liyang Wang, Yu Cheng, Hao Gong, Jiacheng Hu, Xirui Tang, Iris Li
Abstract: The sophistication and diversity of contemporary cyberattacks have rendered the use of proxies, gateways, firewalls, and encrypted tunnels as a standalone defensive strategy inadequate. Consequently, the proactive identification of data anomalies has emerged as a prominent area of research within the field of data security. The majority of extant studies concentrate on sample equilibrium data, with the consequence that the detection effect is not optimal in the context of unbalanced data. In this study, the unsupervised learning method is employed to identify anomalies in dynamic data flows. Initially, multi-dimensional features are extracted from real-time data, and a clustering algorithm is utilised to analyse the patterns of the data. This enables the potential outliers to be automatically identified. By clustering similar data, the model is able to detect data behaviour that deviates significantly from normal traffic without the need for labelled data. The results of the experiments demonstrate that the proposed method exhibits high accuracy in the detection of anomalies across a range of scenarios. Notably, it demonstrates robust and adaptable performance, particularly in the context of unbalanced data.
Authors: Alessio Mascolini, Sebastiano Gaiardelli, Francesco Ponzio, Nicola Dall'Ora, Enrico Macii, Sara Vinco, Santa Di Cataldo, Franco Fummi
Abstract: Detecting complex anomalies on massive amounts of data is a crucial task in Industry 4.0, best addressed by deep learning. However, available solutions are computationally demanding, requiring cloud architectures prone to latency and bandwidth issues. This work presents VARADE, a novel solution implementing a light autoregressive framework based on variational inference, which is best suited for real-time execution on the edge. The proposed approach was validated on a robotic arm, part of a pilot production line, and compared with several state-of-the-art algorithms, obtaining the best trade-off between anomaly detection accuracy, power consumption and inference frequency on two different edge platforms.
Authors: Qinzhuo Wu, Weikai Xu, Wei Liu, Tao Tan, Jianfeng Liu, Ang Li, Jian Luan, Bin Wang, Shuo Shang
Abstract: Recently, mobile AI agents based on VLMs have been gaining increasing attention. These works typically utilize VLM as a foundation, fine-tuning it with instruction-based mobile datasets. However, these VLMs are typically pre-trained on general-domain data, which often results in a lack of fundamental capabilities specific to the mobile domain. Therefore, they may struggle to recognize specific UI elements and understand intra-UI fine-grained information. In addition, the current fine-tuning task focuses on interacting with the most relevant element for the given instruction. These fine-tuned VLMs may still ignore the relationships between UI pages, neglect the roles of elements in page transitions and lack inter-UI understanding. To address issues, we propose a VLM called MobileVLM, which includes two additional pre-training stages to enhance both intra- and inter-UI understanding. We defined four UI-based pre-training tasks, enabling the model to better perceive fine-grained elements and capture page transition actions. To address the lack of mobile pre-training data, we built a large Chinese mobile dataset Mobile3M from scratch, which contains 3 million UI pages, and real-world transition actions, forming a directed graph structure. Experimental results show MobileVLM excels on both our test set and public mobile benchmarks, outperforming existing VLMs.
Authors: Nianqi Li, Siyu Yuan, Jiangjie Chen, Jiaqing Liang, Feng Wei, Zujie Liang, Deqing Yang, Yanghua Xiao
Abstract: Historical analogies, which compare known past events with contemporary but unfamiliar events, are important abilities that help people make decisions and understand the world. However, research in applied history suggests that people have difficulty finding appropriate analogies. And previous studies in the AI community have also overlooked historical analogies. To fill this gap, in this paper, we focus on the historical analogy acquisition task, which aims to acquire analogous historical events for a given event. We explore retrieval and generation methods for acquiring historical analogies based on different large language models (LLMs). Furthermore, we propose a self-reflection method to mitigate hallucinations and stereotypes when LLMs generate historical analogies. Through human evaluations and our specially designed automatic multi-dimensional assessment, we find that LLMs generally have a good potential for historical analogies. And the performance of the models can be further improved by using our self-reflection method.
Authors: Junyu Xue, Yu Zhang, Xudong Wang, Yi Wang, Guoming Tang
Abstract: Non-intrusive load monitoring (NILM), as a key load monitoring technology, can much reduce the deployment cost of traditional power sensors. Previous research has largely focused on developing cloud-exclusive NILM algorithms, which often result in high computation costs and significant service delays. To address these issues, we propose a three-tier framework to enhance the real-world applicability of NILM systems through edge-cloud collaboration. Considering the computational resources available at both the edge and cloud, we implement a lightweight NILM model at the edge and a deep learning based model at the cloud, respectively. In addition to the differential model implementations, we also design a NILM-specific deployment scheme that integrates Gunicorn and NGINX to bridge the gap between theoretical algorithms and practical applications. To verify the effectiveness of the proposed framework, we apply real-world NILM scenario settings and implement the entire process of data acquisition, model training, and system deployment. The results demonstrate that our framework can achieve high decomposition accuracy while significantly reducing the cloud workload and communication overhead under practical considerations.
Authors: Qinzhuo Wu, Wei Liu, Jian Luan, Bin Wang
Abstract: Recently, tool-augmented LLMs have gained increasing attention. Given an instruction, tool-augmented LLMs can interact with various external tools in multiple rounds and provide a final answer. However, previous LLMs were trained on overly detailed instructions, which included API names or parameters, while real users would not explicitly mention these API details. This leads to a gap between trained LLMs and real-world scenarios. In addition, most works ignore whether the interaction process follows the instruction. To address these issues, we constructed a training dataset called MGToolBench, which contains statement and category-level instructions to better reflect real-world scenarios. In addition, we propose ToolPlanner, a two-stage reinforcement learning framework that utilizes path planning and two feedback mechanisms to enhance the LLM's task completion and instruction-following capabilities. Experimental results show that ToolPlanner significantly improves the Match Rate, Pass Rate and Win Rate by 26.8%, 20.2%, and 5.6% compared to the SOTA model. Human evaluation verifies that the multi-granularity instructions can better align with users' usage habits. Our data and code will be released upon acceptance.
Authors: Jiayi Zhang, Chenxin Sun, Yue Gu, Qingyu Zhang, Jiayi Lin, Xiaojiang Du, Chenxiong Qian
Abstract: The gaming industry has experienced substantial growth, but cheating in online games poses a significant threat to the integrity of the gaming experience. Cheating, particularly in first-person shooter (FPS) games, can lead to substantial losses for the game industry. Existing anti-cheat solutions have limitations, such as client-side hardware constraints, security risks, server-side unreliable methods, and both-sides suffer from a lack of comprehensive real-world datasets. To address these limitations, the paper proposes HAWK, a server-side FPS anti-cheat framework for the popular game CS:GO. HAWK utilizes machine learning techniques to mimic human experts' identification process, leverages novel multi-view features, and it is equipped with a well-defined workflow. The authors evaluate HAWK with the first large and real-world datasets containing multiple cheat types and cheating sophistication, and it exhibits promising efficiency and acceptable overheads, shorter ban times compared to the in-use anti-cheat, a significant reduction in manual labor, and the ability to capture cheaters who evaded official inspections.
Authors: Chenxu Yang, Ruipeng Jia, Naibin Gu, Zheng Lin, Siyuan Chen, Chao Pang, Weichong Yin, Yu Sun, Hua Wu, Weiping Wang
Abstract: DPO is an effective preference optimization algorithm. However, the DPO-tuned models tend to overfit on the dispreferred samples, manifested as overly long generations lacking diversity. While recent regularization approaches have endeavored to alleviate this issue by modifying the objective function, they achieved that at the cost of alignment performance degradation. In this paper, we innovatively incorporate regularization from the perspective of weight updating to curb alignment overfitting. Through the pilot experiment, we discovered that there exists a positive correlation between overfitting and the hyperspherical energy fluctuation. Hence, we introduce orthogonal finetuning for DPO via a weight-Rotated Preference Optimization (RoPO) method, which merely conducts rotational and magnitude-stretching updates on the weight parameters to maintain the hyperspherical energy invariant, thereby preserving the knowledge encoded in the angle between neurons. Extensive experiments demonstrate that our model aligns perfectly with human preferences while retaining the original expressive capacity using only 0.0086% of the trainable parameters, suggesting an effective regularization against overfitting. Specifically, RoPO outperforms DPO by up to 10 points on MT-Bench and by up to 2.8 points on AlpacaEval 2, while enhancing the generation diversity by an average of 6 points.
Authors: Aur\'elien Cecille, Stefan Duffner, Franck Davoine, Thibault Neveu, R\'emi Agier
Abstract: Monocular depth estimation has greatly improved in the recent years but models predicting metric depth still struggle to generalize across diverse camera poses and datasets. While recent supervised methods mitigate this issue by leveraging ground prior information at inference, their adaptability to self-supervised settings is limited due to the additional challenge of scale recovery. Addressing this gap, we propose in this paper a novel constraint on ground areas designed specifically for the self-supervised paradigm. This mechanism not only allows to accurately recover the scale but also ensures coherence between the depth prediction and the ground prior. Experimental results show that our method surpasses existing scale recovery techniques on the KITTI benchmark and significantly enhances model generalization capabilities. This improvement can be observed by its more robust performance across diverse camera rotations and its adaptability in zero-shot conditions with previously unseen driving datasets such as DDAD.
Authors: Md. Atiqur Rahman, Nahian Ibn Asad, Md. Mushfiqul Haque Omi, Md. Bakhtiar Hasan, Sabbir Ahmed, Md. Hasanul Kabir
Abstract: Automatic Traffic Sign Recognition is paramount in modern transportation systems, motivating several research endeavors to focus on performance improvement by utilizing large-scale datasets. As the appearance of traffic signs varies across countries, curating large-scale datasets is often impractical; and requires efficient models that can produce satisfactory performance using limited data. In this connection, we present 'FUSED-Net', built-upon Faster RCNN for traffic sign detection, enhanced by Unfrozen Parameters, Pseudo-Support Sets, Embedding Normalization, and Domain Adaptation while reducing data requirement. Unlike traditional approaches, we keep all parameters unfrozen during training, enabling FUSED-Net to learn from limited samples. The generation of a Pseudo-Support Set through data augmentation further enhances performance by compensating for the scarcity of target domain data. Additionally, Embedding Normalization is incorporated to reduce intra-class variance, standardizing feature representation. Domain Adaptation, achieved by pre-training on a diverse traffic sign dataset distinct from the target domain, improves model generalization. Evaluating FUSED-Net on the BDTSD dataset, we achieved 2.4x, 2.2x, 1.5x, and 1.3x improvements of mAP in 1-shot, 3-shot, 5-shot, and 10-shot scenarios, respectively compared to the state-of-the-art Few-Shot Object Detection (FSOD) models. Additionally, we outperform state-of-the-art works on the cross-domain FSOD benchmark under several scenarios.
Authors: Louis Mozart Kamdem Teyou, Caglar Demir, Axel-Cyrille Ngonga Ngomo
Abstract: We introduce a novel embedding method diverging from conventional approaches by operating within function spaces of finite dimension rather than finite vector space, thus departing significantly from standard knowledge graph embedding techniques. Initially employing polynomial functions to compute embeddings, we progress to more intricate representations using neural networks with varying layer complexities. We argue that employing functions for embedding computation enhances expressiveness and allows for more degrees of freedom, enabling operations such as composition, derivatives and primitive of entities representation. Additionally, we meticulously outline the step-by-step construction of our approach and provide code for reproducibility, thereby facilitating further exploration and application in the field.
Authors: Xueluan Gong, Mingzhe Li, Yilin Zhang, Fengyuan Ran, Chen Chen, Yanjiao Chen, Qian Wang, Kwok-Yan Lam
Abstract: Large Language Models (LLMs) have excelled in various tasks but are still vulnerable to jailbreaking attacks, where attackers create jailbreak prompts to mislead the model to produce harmful or offensive content. Current jailbreak methods either rely heavily on manually crafted templates, which pose challenges in scalability and adaptability, or struggle to generate semantically coherent prompts, making them easy to detect. Additionally, most existing approaches involve lengthy prompts, leading to higher query costs.In this paper, to remedy these challenges, we introduce a novel jailbreaking attack framework, which is an automated, black-box jailbreaking attack framework that adapts the black-box fuzz testing approach with a series of customized designs. Instead of relying on manually crafted templates, our method starts with an empty seed pool, removing the need to search for any related jailbreaking templates. We also develop three novel question-dependent mutation strategies using an LLM helper to generate prompts that maintain semantic coherence while significantly reducing their length. Additionally, we implement a two-level judge module to accurately detect genuine successful jailbreaks. We evaluated our method on 7 representative LLMs and compared it with 5 state-of-the-art jailbreaking attack strategies. For proprietary LLM APIs, such as GPT-3.5 turbo, GPT-4, and Gemini-Pro, our method achieves attack success rates of over 90%, 80%, and 74%, respectively, exceeding existing baselines by more than 60%. Additionally, our method can maintain high semantic coherence while significantly reducing the length of jailbreak prompts. When targeting GPT-4, our method can achieve over 78\% attack success rate even with 100 tokens. Moreover, our method demonstrates transferability and is robust to state-of-the-art defenses. We will open-source our codes upon publication.
Authors: Grigory Yaremenko, Dmitrii Dobriborsci, Roman Zashchitin, Ruben Contreras Maestre, Ngoc Quoc Huy Hoang, Pavel Osinenko
Abstract: Reinforcement Learning (RL) has been shown to be effective and convenient for a number of tasks in robotics. However, it requires the exploration of a sufficiently large number of state-action pairs, many of which may be unsafe or unimportant. For instance, online model-free learning can be hazardous and inefficient in the absence of guarantees that a certain set of desired states will be reached during an episode. An increasingly common approach to address safety involves the addition of a shielding system that constrains the RL actions to a safe set of actions. In turn, a difficulty for such frameworks is how to effectively couple RL with the shielding system to make sure the exploration is not excessively restricted. This work presents a novel safe model-free RL agent called Critic As Lyapunov Function (CALF) and showcases how CALF can be used to improve upon control baselines in robotics in an efficient and convenient fashion while ensuring guarantees of stable goal reaching. The latter is a crucial part of safety, as seen generally. With CALF all state-action pairs remain explorable and yet reaching of desired goal states is formally guaranteed. Formal analysis is provided that shows the goal stabilization-ensuring properties of CALF and a set of real-world and numerical experiments with a non-holonomic wheeled mobile robot (WMR) TurtleBot3 Burger confirmed the superiority of CALF over such a well-established RL agent as proximal policy optimization (PPO), and a modified version of SARSA in a few-episode setting in terms of attained total cost.
Authors: Yongxin Deng, Xiaoyu Tan, Xihe Qiu, Yaochu Jin
Abstract: Reinforcement learning methods have been used to optimize long-term user engagement in recommendation systems. However, existing reinforcement learning-based recommendation systems do not fully exploit the relevance of individual user behavior across different platforms. One potential solution is to aggregate data from various platforms in a centralized location and use the aggregated data for training. However, this approach raises economic and legal concerns, including increased communication costs and potential threats to user privacy. To address these challenges, we propose \textbf{FedSlate}, a federated reinforcement learning recommendation algorithm that effectively utilizes information that is prohibited from being shared at a legal level. We employ the SlateQ algorithm to assist FedSlate in learning users' long-term behavior and evaluating the value of recommended content. We extend the existing application scope of recommendation systems from single-user single-platform to single-user multi-platform and address cross-platform learning challenges by introducing federated learning. We use RecSim to construct a simulation environment for evaluating FedSlate and compare its performance with state-of-the-art benchmark recommendation models. Experimental results demonstrate the superior effects of FedSlate over baseline methods in various environmental settings, and FedSlate facilitates the learning of recommendation strategies in scenarios where baseline methods are completely inapplicable. Code is available at \textit{https://github.com/TianYaDY/FedSlate}.
Authors: Ahjol Senbi, Tianyu Huang, Fei Lyu, Qing Li, Yuhui Tao, Wei Shao, Qiang Chen, Chengyan Wang, Shuo Wang, Tao Zhou, Yizhe Zhang
Abstract: We are interested in building a ground-truth-free evaluation model to assess the quality of segmentations produced by SAM (Segment Anything Model) and its variants in medical images. This model estimates segmentation quality scores by comparing input images with their corresponding segmentation maps. Building on prior research, we frame this as a regression problem within a supervised learning framework, using Dice scores (and optionally other metrics) to compute the training loss. The model is trained using a large collection of public datasets of medical images with segmentation predictions from SAM and its variants. We name this model EvanySeg (Evaluation of Any Segmentation in Medical Images). Our exploration of convolution-based models (e.g., ResNet) and transformer-based models (e.g., ViT) revealed that ViT offers superior performance for EvanySeg. This model can be employed for various tasks, including: (1) identifying poorly segmented samples by detecting low-percentile segmentation quality scores; (2) benchmark segmentation models without ground truth by averaging scores across test samples; (3) alerting human experts during human-AI collaboration by applying a threshold within the score space; and (4) selecting the best segmentation prediction for each test sample at test time when multiple segmentation models are available, by choosing the prediction with the highest score. Models and code will be made available at https://github.com/ahjolsenbics/EvanySeg.
Authors: Shilong Yang, Chulong Zhang, Qi Zang, Juan Yu, Liang Zeng, Xiao Luo, Yexuan Xing, Xin Pan, Qi Li, Xiaokun Liang, Yaoqin Xie
Abstract: Breast cancer has long posed a significant threat to women's health, making early screening crucial for mitigating its impact. However, mammography, the preferred method for early screening, faces limitations such as the burden of double reading by radiologists, challenges in widespread adoption in remote and underdeveloped areas, and obstacles in intelligent early screening development due to data constraints. To address these challenges, we propose a weakly supervised multi-view mammography early screening model for breast cancer based on context clustering. Context clustering, a feature extraction structure that is neither CNN nor transformer, combined with multi-view learning for information complementation, presents a promising approach. The weak supervision design specifically addresses data limitations. Our model achieves state-of-the-art performance with fewer parameters on two public datasets, with an AUC of 0.828 on the Vindr-Mammo dataset and 0.805 on the CBIS-DDSM dataset. Our model shows potential in reducing the burden on doctors and increasing the feasibility of breast cancer screening for women in underdeveloped regions.
Authors: Bin Hong, Jinze Wu, Jiayu Liu, Liang Ding, Jing Sha, Kai Zhang, Shijin Wang, Zhenya Huang
Abstract: In recent years, the breakthrough of Large Language Models (LLMs) offers new ideas for achieving universal methods on graph data. The common practice of converting graphs into natural language for LLMs, which refers to graph flattening, exhibits good generalizability and interpretability. However, the poor organization of the textual format results in poor performance in long-distance scenario understanding. Inspired by human cognitive reasoning habits, we propose a novel method for graph flattening to fit LLMs, termed as End-to-End DAG-Path prompting (EEDP). Experiments on real-world datasets show that EEDP enhances the reasoning performance of LLMs in long-distance scenarios while maintaining excellent performance in short-distance scenarios, demonstrating good robustness in the face of distance variations.
Authors: Yannis Bendi-Ouis, Dan Dutarte, Xavier Hinaut
Abstract: Since the release of ChatGPT in November 2023, large language models (LLMs) have seen considerable success, including in the open-source community, with many open-weight models available. However, the requirements to deploy such a service are often unknown and difficult to evaluate in advance. To facilitate this process, we conducted numerous tests at the Centre Inria de l'Universit\'e de Bordeaux. In this article, we propose a comparison of the performance of several models of different sizes (mainly Mistral and LLaMa) depending on the available GPUs, using vLLM, a Python library designed to optimize the inference of these models. Our results provide valuable information for private and public groups wishing to deploy LLMs, allowing them to evaluate the performance of different models based on their available hardware. This study thus contributes to facilitating the adoption and use of these large language models in various application domains.
Authors: Sangyeon Cho, Jangyeong Jeon, Dongjoon Lee, Changhee Lee, Junyeong Kim
Abstract: The use of pre-trained language models fine-tuned to address specific downstream tasks is a common approach in natural language processing (NLP). However, acquiring domain-specific knowledge via fine-tuning is challenging. Traditional methods involve pretraining language models using vast amounts of domain-specific data before fine-tuning for particular tasks. This study investigates emergency/non-emergency classification tasks based on electronic medical record (EMR) data obtained from pediatric emergency departments (PEDs) in Korea. Our findings reveal that existing domain-specific pre-trained language models underperform compared to general language models in handling N-lingual free-text data characteristics of non-English-speaking regions. To address these limitations, we propose a domain knowledge transfer methodology that leverages knowledge distillation to infuse general language models with domain-specific knowledge via fine-tuning. This study demonstrates the effective transfer of specialized knowledge between models by defining a general language model as the student model and a domain-specific pre-trained model as the teacher model. In particular, we address the complexities of EMR data obtained from PEDs in non-English-speaking regions, such as Korea, and demonstrate that the proposed method enhances classification performance in such contexts. The proposed methodology not only outperforms baseline models on Korean PED EMR data, but also promises broader applicability in various professional and technical domains. In future works, we intend to extend this methodology to include diverse non-English-speaking regions and address additional downstream tasks, with the aim of developing advanced model architectures using state-of-the-art KD techniques. The code is available in https://github.com/JoSangYeon/DSG-KD.
Authors: Zixuan Wang, Bo Yu, Junzhe Zhao, Wenhao Sun, Sai Hou, Shuai Liang, Xing Hu, Yinhe Han, Yiming Gan
Abstract: Embodied AI agents responsible for executing interconnected, long-sequence household tasks often face difficulties with in-context memory, leading to inefficiencies and errors in task execution. To address this issue, we introduce KARMA, an innovative memory system that integrates long-term and short-term memory modules, enhancing large language models (LLMs) for planning in embodied agents through memory-augmented prompting. KARMA distinguishes between long-term and short-term memory, with long-term memory capturing comprehensive 3D scene graphs as representations of the environment, while short-term memory dynamically records changes in objects' positions and states. This dual-memory structure allows agents to retrieve relevant past scene experiences, thereby improving the accuracy and efficiency of task planning. Short-term memory employs strategies for effective and adaptive memory replacement, ensuring the retention of critical information while discarding less pertinent data. Compared to state-of-the-art embodied agents enhanced with memory, our memory-augmented embodied AI agent improves success rates by 1.3x and 2.3x in Composite Tasks and Complex Tasks within the AI2-THOR simulator, respectively, and enhances task execution efficiency by 3.4x and 62.7x. Furthermore, we demonstrate that KARMA's plug-and-play capability allows for seamless deployment on real-world robotic systems, such as mobile manipulation platforms.Through this plug-and-play memory system, KARMA significantly enhances the ability of embodied agents to generate coherent and contextually appropriate plans, making the execution of complex household tasks more efficient. The experimental videos from the work can be found at https://youtu.be/4BT7fnw9ehs.
Authors: Siyun Zhao, Yuqing Yang, Zilong Wang, Zhiyuan He, Luna K. Qiu, Lili Qiu
Abstract: Large language models (LLMs) augmented with external data have demonstrated remarkable capabilities in completing real-world tasks. Techniques for integrating external data into LLMs, such as Retrieval-Augmented Generation (RAG) and fine-tuning, are gaining increasing attention and widespread application. Nonetheless, the effective deployment of data-augmented LLMs across various specialized fields presents substantial challenges. These challenges encompass a wide range of issues, from retrieving relevant data and accurately interpreting user intent to fully harnessing the reasoning capabilities of LLMs for complex tasks. We believe that there is no one-size-fits-all solution for data-augmented LLM applications. In practice, underperformance often arises from a failure to correctly identify the core focus of a task or because the task inherently requires a blend of multiple capabilities that must be disentangled for better resolution. In this survey, we propose a RAG task categorization method, classifying user queries into four levels based on the type of external data required and primary focus of the task: explicit fact queries, implicit fact queries, interpretable rationale queries, and hidden rationale queries. We define these levels of queries, provide relevant datasets, and summarize the key challenges and most effective techniques for addressing these challenges. Finally, we discuss three main forms of integrating external data into LLMs: context, small model, and fine-tuning, highlighting their respective strengths, limitations, and the types of problems they are suited to solve. This work aims to help readers thoroughly understand and decompose the data requirements and key bottlenecks in building LLM applications, offering solutions to the different challenges and serving as a guide to systematically developing such applications.
Authors: Keqin Li, Jiajing Chen, Denzhi Yu, Tao Dajun, Xinyu Qiu, Lian Jieting, Sun Baiwei, Zhang Shengyuan, Zhenyu Wan, Ran Ji, Bo Hong, Fanghao Ni
Abstract: At present, in most warehouse environments, the accumulation of goods is complex, and the management personnel in the control of goods at the same time with the warehouse mobile robot trajectory interaction, the traditional mobile robot can not be very good on the goods and pedestrians to feed back the correct obstacle avoidance strategy, in order to control the mobile robot in the warehouse environment efficiently and friendly to complete the obstacle avoidance task, this paper proposes a deep reinforcement learning based on the warehouse environment, the mobile robot obstacle avoidance Algorithm. Firstly, for the insufficient learning ability of the value function network in the deep reinforcement learning algorithm, the value function network is improved based on the pedestrian interaction, the interaction information between pedestrians is extracted through the pedestrian angle grid, and the temporal features of individual pedestrians are extracted through the attention mechanism, so that we can learn to obtain the relative importance of the current state and the historical trajectory state as well as the joint impact on the robot's obstacle avoidance strategy, which provides an opportunity for the learning of multi-layer perceptual machines afterwards. Secondly, the reward function of reinforcement learning is designed based on the spatial behaviour of pedestrians, and the robot is punished for the state where the angle changes too much, so as to achieve the requirement of comfortable obstacle avoidance; Finally, the feasibility and effectiveness of the deep reinforcement learning-based mobile robot obstacle avoidance algorithm in the warehouse environment in the complex environment of the warehouse are verified through simulation experiments.
Authors: Devon Jarvis, Richard Klein, Benjamin Rosman, Andrew M. Saxe
Abstract: A number of machine learning models have been proposed with the goal of achieving systematic generalization: the ability to reason about new situations by combining aspects of previous experiences. These models leverage compositional architectures which aim to learn specialized modules dedicated to structures in a task that can be composed to solve novel problems with similar structures. While the compositionality of these architectures is guaranteed by design, the modules specializing is not. Here we theoretically study the ability of network modules to specialize to useful structures in a dataset and achieve systematic generalization. To this end we introduce a minimal space of datasets motivated by practical systematic generalization benchmarks. From this space of datasets we present a mathematical definition of systematicity and study the learning dynamics of linear neural modules when solving components of the task. Our results shed light on the difficulty of module specialization, what is required for modules to successfully specialize, and the necessity of modular architectures to achieve systematicity. Finally, we confirm that the theoretical results in our tractable setting generalize to more complex datasets and non-linear architectures.
Authors: Jiashuo Li, Shaokun Wang, Bo Qian, Yuhang He, Xing Wei, Yihong Gong
Abstract: Non-exemplar class Incremental Learning (NECIL) enables models to continuously acquire new classes without retraining from scratch and storing old task exemplars, addressing privacy and storage issues. However, the absence of data from earlier tasks exacerbates the challenge of catastrophic forgetting in NECIL. In this paper, we propose a novel framework called Dynamic Integration of task-specific Adapters (DIA), which comprises two key components: Task-Specific Adapter Integration (TSAI) and Patch-Level Model Alignment. TSAI boosts compositionality through a patch-level adapter integration strategy, which provides a more flexible compositional solution while maintaining low computation costs. Patch-Level Model Alignment maintains feature consistency and accurate decision boundaries via two specialized mechanisms: Patch-Level Distillation Loss (PDL) and Patch-Level Feature Reconstruction method (PFR). Specifically, the PDL preserves feature-level consistency between successive models by implementing a distillation loss based on the contributions of patch tokens to new class learning. The PFR facilitates accurate classifier alignment by reconstructing old class features from previous tasks that adapt to new task knowledge. Extensive experiments validate the effectiveness of our DIA, revealing significant improvements on benchmark datasets in the NECIL setting, maintaining an optimal balance between computational complexity and accuracy. The full code implementation will be made publicly available upon the publication of this paper.
Authors: Minseung Lee, Seokha Moon, Seung Joon Lee, Jinkyu Kim
Abstract: Accurately detecting objects at long distances remains a critical challenge in 3D object detection when relying solely on LiDAR sensors due to the inherent limitations of data sparsity. To address this issue, we propose the LiDAR-Camera Augmentation Network (LCANet), a novel framework that reconstructs LiDAR point cloud data by fusing 2D image features, which contain rich semantic information, generating additional points to improve detection accuracy. LCANet fuses data from LiDAR sensors and cameras by projecting image features into the 3D space, integrating semantic information into the point cloud data. This fused data is then encoded to produce 3D features that contain both semantic and spatial information, which are further refined to reconstruct final points before bounding box prediction. This fusion effectively compensates for LiDAR's weakness in detecting objects at long distances, which are often represented by sparse points. Additionally, due to the sparsity of many objects in the original dataset, which makes effective supervision for point generation challenging, we employ a point cloud completion network to create a complete point cloud dataset that supervises the generation of dense point clouds in our network. Extensive experiments on the KITTI and Waymo datasets demonstrate that LCANet significantly outperforms existing models, particularly in detecting sparse and distant objects.
Authors: Anthony Sicilia, Malihe Alikhani
Abstract: Typically, when evaluating Theory of Mind, we consider the beliefs of others to be binary: held or not held. But what if someone is unsure about their own beliefs? How can we quantify this uncertainty? We propose a new suite of tasks, challenging language models (LMs) to model the uncertainty of others in dialogue. We design these tasks around conversation forecasting, wherein an agent forecasts an unobserved outcome to a conversation. Uniquely, we view interlocutors themselves as forecasters, asking an LM to predict the uncertainty of the interlocutors (a probability). We experiment with re-scaling methods, variance reduction strategies, and demographic context, for this regression task, conducting experiments on three dialogue corpora (social, negotiation, task-oriented) with eight LMs. While LMs can explain up to 7% variance in the uncertainty of others, we highlight the difficulty of the tasks and room for future work, especially in practical applications, like anticipating ``false
Authors: Georgios Papasotiropoulos, Seyedeh Zeinab Pishbin, Oskar Skibski, Piotr Skowron, Tomasz W\k{a}s
Abstract: In participatory budgeting (PB), voters decide through voting which subset of projects to fund within a given budget. Proportionality in the context of PB is crucial to ensure equal treatment of all groups of voters. However, pure proportional rules can sometimes lead to suboptimal outcomes. We introduce the Method of Equal Shares with Bounded Overspending (BOS Equal Shares), a robust variant of Equal Shares that balances proportionality and efficiency. BOS Equal Shares addresses inefficiencies inherent in strict proportionality guarantees yet still provides good proportionality similar to the original Method of Equal Shares. In the course of the analysis, we also discuss a fractional variant of the method which allows for partial funding of projects.
Authors: Sijia Du, Chengfeng Zhou, Suncheng Xiang, Jianwei Xu, Dahong Qian
Abstract: Objective: Depth estimation is crucial for endoscopic navigation and manipulation, but obtaining ground-truth depth maps in real clinical scenarios, such as the colon, is challenging. This study aims to develop a robust framework that generalizes well to real colonoscopy images, overcoming challenges like non-Lambertian surface reflection and diverse data distributions. Methods: We propose a framework combining a convolutional neural network (CNN) for capturing local features and a Transformer for capturing global information. An uncertainty-based fusion block was designed to enhance generalization by identifying complementary contributions from the CNN and Transformer branches. The network can be trained with simulated datasets and generalize directly to unseen clinical data without any fine-tuning. Results: Our method is validated on multiple datasets and demonstrates an excellent generalization ability across various datasets and anatomical structures. Furthermore, qualitative analysis in real clinical scenarios confirmed the robustness of the proposed method. Conclusion: The integration of local and global features through the CNN-Transformer architecture, along with the uncertainty-based fusion block, improves depth estimation performance and generalization in both simulated and real-world endoscopic environments. Significance: This study offers a novel approach to estimate depth maps for endoscopy images despite the complex conditions in clinic, serving as a foundation for endoscopic automatic navigation and other clinical tasks, such as polyp detection and segmentation.
Authors: Shashank Rajput, Ying Sheng, Sean Owen, Vitaliy Chiley
Abstract: The size of the key-value (KV) cache plays a critical role in determining both the maximum context length and the number of concurrent requests supported during inference in modern language models. The KV cache size grows proportionally with the number of attention heads and the tokens processed, leading to increased memory consumption and slower inference for long inputs. In this work, we explore the use of MixAttention, a model architecture modification closely related to a blog published by Character.AI. MixAttention combines sliding window attention, where only a small subset of recent tokens is stored in the KV cache, with KV cache sharing across layers. Our experiments demonstrate that MixAttention significantly reduces memory usage and improves inference speed without sacrificing model performance in both short and long-context tasks. We also explore various configurations of this architecture, identifying those that maintain quality across evaluation metrics while optimizing resource efficiency.
Authors: Svea Marie Meyer, Philipp Weidel, Philipp Plank, Leobardo Campos-Macias, Sumit Bam Shrestha, Philipp Stratmann, Mathis Richter
Abstract: Deep State-Space Models (SSM) demonstrate state-of-the art performance on long-range sequence modeling tasks. While the recurrent structure of SSMs can be efficiently implemented as a convolution or as a parallel scan during training, recurrent token-by-token processing cannot currently be implemented efficiently on GPUs. Here, we demonstrate efficient token-by-token inference of the SSM S4D on Intel's Loihi 2 state-of-the-art neuromorphic processor. We compare this first ever neuromorphic-hardware implementation of an SSM on sMNIST, psMNIST, and sCIFAR to a recurrent and a convolutional implementation of S4D on Jetson Orin Nano (Jetson). While we find Jetson to perform better in an offline sample-by-sample based batched processing mode, Loihi 2 outperforms during token-by-token based processing, where it consumes 1000 times less energy with a 75 times lower latency and a 75 times higher throughput compared to the recurrent implementation of S4D on Jetson. This opens up new avenues towards efficient real-time streaming applications of SSMs.
Authors: Mohammad Amin Roshani, Xiangyu Zhou, Yao Qiang, Srinivasan Suresh, Steve Hicks, Usha Sethuraman, Dongxiao Zhu
Abstract: Large language models (LLMs) have shown remarkable capabilities in various natural language tasks and are increasingly being applied in healthcare domains. This work demonstrates a new LLM-powered disease risk assessment approach via streaming human-AI conversation, eliminating the need for programming required by traditional machine learning approaches. In a COVID-19 severity risk assessment case study, we fine-tune pre-trained generative LLMs (e.g., Llama2-7b and Flan-t5-xl) using a few shots of natural language examples, comparing their performance with traditional classifiers (i.e., Logistic Regression, XGBoost, Random Forest) that are trained de novo using tabular data across various experimental settings. We develop a mobile application that uses these fine-tuned LLMs as its generative AI (GenAI) core to facilitate real-time interaction between clinicians and patients, providing no-code risk assessment through conversational interfaces. This integration not only allows for the use of streaming Questions and Answers (QA) as inputs but also offers personalized feature importance analysis derived from the LLM's attention layers, enhancing the interpretability of risk assessments. By achieving high Area Under the Curve (AUC) scores with a limited number of fine-tuning samples, our results demonstrate the potential of generative LLMs to outperform discriminative classification methods in low-data regimes, highlighting their real-world adaptability and effectiveness. This work aims to fill the existing gap in leveraging generative LLMs for interactive no-code risk assessment and to encourage further research in this emerging field.
Authors: Saptarshi Saha, Utpal Garain
Abstract: This paper introduces a simple extension of mixup (Zhang et al., 2018) data augmentation to enhance generalization in visual recognition tasks. Unlike the vanilla mixup method, which blends entire images, our approach focuses on combining regions from multiple images.
Authors: Swathi Shree Narashiman, Nitin Chandrachoodan
Abstract: Data compression continues to evolve, with traditional information theory methods being widely used for compressing text, images, and videos. Recently, there has been growing interest in leveraging Generative AI for predictive compression techniques. This paper introduces a lossless text compression approach using a Large Language Model (LLM). The method involves two key steps: first, prediction using a dense neural network architecture, such as a transformer block; second, compressing the predicted ranks with standard compression algorithms like Adaptive Huffman, LZ77, or Gzip. Extensive analysis and benchmarking against conventional information-theoretic baselines demonstrate that neural compression offers improved performance.
Authors: Ga\"etan Caillaut, Raheel Qader, Mariam Nakhl\'e, Jingshu Liu, Jean-Gabriel Barth\'elemy
Abstract: Recent studies have showcased remarkable capabilities of decoder-only models in many NLP tasks, including translation. Yet, the machine translation field has been largely dominated by encoder-decoder models based on the Transformer architecture. As a consequence, scaling laws of encoder-decoder models for neural machine translation have already been well studied, but decoder-only models have received less attention. This work explores the scaling laws of decoder-only models on the multilingual and multidomain translation task. We trained a collection of six decoder-only models, ranging from 70M to 7B parameters, on a sentence-level, multilingual and multidomain dataset. We conducted a series of experiments showing that the loss of decoder-only models can be estimated using a scaling law similar to the one discovered for large language models, but we also show that this scaling law has difficulties to generalize to too large models or to a different data distribution. We also study different scaling methods and show that scaling the depth and the width of a model lead to similar test loss improvements, but with different impact on the model's efficiency.
Authors: Siddharth Betala, Ishan Chokshi
Abstract: In this paper, we describe our system under the team name Brotherhood for the English-to-Lowres Multi-Modal Translation Task. We participate in the multi-modal translation tasks for English-Hindi, English-Hausa, English-Bengali, and English-Malayalam language pairs. We present a method leveraging multi-modal Large Language Models (LLMs), specifically GPT-4o and Claude 3.5 Sonnet, to enhance cross-lingual image captioning without traditional training or fine-tuning. Our approach utilizes instruction-tuned prompting to generate rich, contextual conversations about cropped images, using their English captions as additional context. These synthetic conversations are then translated into the target languages. Finally, we employ a weighted prompting strategy, balancing the original English caption with the translated conversation to generate captions in the target language. This method achieved competitive results, scoring 37.90 BLEU on the English-Hindi Challenge Set and ranking first and second for English-Hausa on the Challenge and Evaluation Leaderboards, respectively. We conduct additional experiments on a subset of 250 images, exploring the trade-offs between BLEU scores and semantic similarity across various weighting schemes.
Authors: Sean Kim, Raja Mazumder
Abstract: The exponential growth in computational power and accessibility has transformed the complexity and scale of bioinformatics research, necessitating standardized documentation for transparency, reproducibility, and regulatory compliance. The IEEE BioCompute Object (BCO) standard addresses this need but faces adoption challenges due to the overhead of creating compliant documentation, especially for legacy research. This paper presents a novel approach to automate the creation of BCOs from scientific papers using Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs). We describe the development of the BCO assistant tool that leverages RAG to extract relevant information from source papers and associated code repositories, addressing key challenges such as LLM hallucination and long-context understanding. The implementation incorporates optimized retrieval processes, including a two-pass retrieval with re-ranking, and employs carefully engineered prompts for each BCO domain. We discuss the tool's architecture, extensibility, and evaluation methods, including automated and manual assessment approaches. The BCO assistant demonstrates the potential to significantly reduce the time and effort required for retroactive documentation of bioinformatics research while maintaining compliance with the standard. This approach opens avenues for AI-assisted scientific documentation and knowledge extraction from publications thereby enhancing scientific reproducibility. The BCO assistant tool and documentation is available at https://biocompute-objects.github.io/bco-rag/.
Authors: Kunyao Lan, Bingui Jin, Zichen Zhu, Siyuan Chen, Shu Zhang, Kenny Q. Zhu, Mengyue Wu
Abstract: Mental health issues, particularly depressive disorders, present significant challenges in contemporary society, necessitating the development of effective automated diagnostic methods. This paper introduces the Agent Mental Clinic (AMC), a self-improving conversational agent system designed to enhance depression diagnosis through simulated dialogues between patient and psychiatrist agents. To enhance the dialogue quality and diagnosis accuracy, we design a psychiatrist agent consisting of a tertiary memory structure, a dialogue control and reflect plugin that acts as ``supervisor'' and a memory sampling module, fully leveraging the skills reflected by the psychiatrist agent, achieving great accuracy on depression risk and suicide risk diagnosis via conversation. Experiment results on datasets collected in real-life scenarios demonstrate that the system, simulating the procedure of training psychiatrists, can be a promising optimization method for aligning LLMs with real-life distribution in specific domains without modifying the weights of LLMs, even when only a few representative labeled cases are available.
Authors: Hongyi Wang, Xiuju Du, Jing Liu, Shuyi Ouyang, Yen-Wei Chen, Lanfen Lin
Abstract: The advancement of Spatial Transcriptomics (ST) has facilitated the spatially-aware profiling of gene expressions based on histopathology images. Although ST data offers valuable insights into the micro-environment of tumors, its acquisition cost remains expensive. Therefore, directly predicting the ST expressions from digital pathology images is desired. Current methods usually adopt existing regression backbones along with patch-sampling for this task, which ignores the inherent multi-scale information embedded in the pyramidal data structure of digital pathology images, and wastes the inter-spot visual information crucial for accurate gene expression prediction. To address these limitations, we propose M2OST, a many-to-one regression Transformer that can accommodate the hierarchical structure of the pathology images via a decoupled multi-scale feature extractor. Unlike traditional models that are trained with one-to-one image-label pairs, M2OST uses multiple images from different levels of the digital pathology image to jointly predict the gene expressions in their common corresponding spot. Built upon our many-to-one scheme, M2OST can be easily scaled to fit different numbers of inputs, and its network structure inherently incorporates nearby inter-spot features, enhancing regression performance. We have tested M2OST on three public ST datasets and the experimental results show that M2OST can achieve state-of-the-art performance with fewer parameters and floating-point operations (FLOPs). The code will be released upon acceptance.
Authors: Daniel Honerkamp, Harsh Mahesheka, Jan Ole von Hartz, Tim Welschehold, Abhinav Valada
Abstract: Demonstration data plays a key role in learning complex behaviors and training robotic foundation models. While effective control interfaces exist for static manipulators, data collection remains cumbersome and time intensive for mobile manipulators due to their large number of degrees of freedom. While specialized hardware, avatars, or motion tracking can enable whole-body control, these approaches are either expensive, robot-specific, or suffer from the embodiment mismatch between robot and human demonstrator. In this work, we present MoMa-Teleop, a novel teleoperation method that delegates the base motions to a reinforcement learning agent, leaving the operator to focus fully on the task-relevant end-effector motions. This enables whole-body teleoperation of mobile manipulators with zero additional hardware or setup costs via standard interfaces such as joysticks or hand guidance. Moreover, the operator is not bound to a tracked workspace and can move freely with the robot over spatially extended tasks. We demonstrate that our approach results in a significant reduction in task completion time across a variety of robots and tasks. As the generated data covers diverse whole-body motions without embodiment mismatch, it enables efficient imitation learning. By focusing on task-specific end-effector motions, our approach learns skills that transfer to unseen settings, such as new obstacles or changed object positions, from as little as five demonstrations. We make code and videos available at http://moma-teleop.cs.uni-freiburg.de.
Authors: Agniv Sharma, Jonas Geiping
Abstract: Transformers are widely used across various applications, many of which yield sparse or partially filled attention matrices. Examples include attention masks designed to reduce the quadratic complexity of attention, sequence packing techniques, and recent innovations like tree masking for fast validation in MEDUSA. Despite the inherent sparsity in these matrices, the state-of-the-art algorithm Flash Attention still processes them with quadratic complexity as though they were dense. In this paper, we introduce \textbf{Binary Block Masking}, a highly efficient modification that enhances Flash Attention by making it mask-aware. We further propose two optimizations: one tailored for masks with contiguous non-zero patterns and another for extremely sparse masks. Our experiments on attention masks derived from real-world scenarios demonstrate up to a 9x runtime improvement. The implementation will be publicly released to foster further research and application.
Authors: Jiaxing Li, Zihan Chen, Kai Fong Ernest Chong, Bikramjit Das, Tony Q. S. Quek, Howard H. Yang
Abstract: Leveraging over-the-air computations for model aggregation is an effective approach to cope with the communication bottleneck in federated edge learning. By exploiting the superposition properties of multi-access channels, this approach facilitates an integrated design of communication and computation, thereby enhancing system privacy while reducing implementation costs. However, the inherent electromagnetic interference in radio channels often exhibits heavy-tailed distributions, giving rise to exceptionally strong noise in globally aggregated gradients that can significantly deteriorate the training performance. To address this issue, we propose a novel gradient clipping method, termed Median Anchored Clipping (MAC), to combat the detrimental effects of heavy-tailed noise. We also derive analytical expressions for the convergence rate of model training with analog over-the-air federated learning under MAC, which quantitatively demonstrates the effect of MAC on training performance. Extensive experimental results show that the proposed MAC algorithm effectively mitigates the impact of heavy-tailed noise, hence substantially enhancing system robustness.
Authors: Tuan-Hung Vu, Eduardo Valle, Andrei Bursuc, Tommie Kerssies, Daan de Geus, Gijs Dubbelman, Long Qian, Bingke Zhu, Yingying Chen, Ming Tang, Jinqiao Wang, Tom\'a\v{s} Voj\'i\v{r}, Jan \v{S}ochman, Ji\v{r}\'i Matas, Michael Smith, Frank Ferrie, Shamik Basu, Christos Sakaridis, Luc Van Gool
Abstract: We propose the unified BRAVO challenge to benchmark the reliability of semantic segmentation models under realistic perturbations and unknown out-of-distribution (OOD) scenarios. We define two categories of reliability: (1) semantic reliability, which reflects the model's accuracy and calibration when exposed to various perturbations; and (2) OOD reliability, which measures the model's ability to detect object classes that are unknown during training. The challenge attracted nearly 100 submissions from international teams representing notable research institutions. The results reveal interesting insights into the importance of large-scale pre-training and minimal architectural design in developing robust and reliable semantic segmentation models.
Authors: Weiping Yu, Siqiang Luo, Zihao Yu, Gao Cong
Abstract: We use machine learning to optimize LSM-tree structure, aiming to reduce the cost of processing various read/write operations. We introduce a new approach Camal, which boasts the following features: (1) ML-Aided: Camal is the first attempt to apply active learning to tune LSM-tree based key-value stores. The learning process is coupled with traditional cost models to improve the training process; (2) Decoupled Active Learning: backed by rigorous analysis, Camal adopts active learning paradigm based on a decoupled tuning of each parameter, which further accelerates the learning process; (3) Easy Extrapolation: Camal adopts an effective mechanism to incrementally update the model with the growth of the data size; (4) Dynamic Mode: Camal is able to tune LSM-tree online under dynamically changing workloads; (5) Significant System Improvement: By integrating Camal into a full system RocksDB, the system performance improves by 28% on average and up to 8x compared to a state-of-the-art RocksDB design.
Authors: Kehui Liu, Zixin Tang, Dong Wang, Zhigang Wang, Bin Zhao, Xuelong Li
Abstract: Leveraging the powerful reasoning capabilities of large language models (LLMs), recent LLM-based robot task planning methods yield promising results. However, they mainly focus on single or multiple homogeneous robots on simple tasks. Practically, complex long-horizon tasks always require collaborations among multiple heterogeneous robots especially with more complex action spaces, which makes these tasks more challenging. To this end, we propose COHERENT, a novel LLM-based task planning framework for collaboration of heterogeneous multi-robot systems including quadrotors, robotic dogs, and robotic arms. Specifically, a Proposal-Execution-Feedback-Adjustment (PEFA) mechanism is designed to decompose and assign actions for individual robots, where a centralized task assigner makes a task planning proposal to decompose the complex task into subtasks, and then assigns subtasks to robot executors. Each robot executor selects a feasible action to implement the assigned subtask and reports self-reflection feedback to the task assigner for plan adjustment. The PEFA loops until the task is completed. Moreover, we create a challenging heterogeneous multi-robot task planning benchmark encompassing 100 complex long-horizon tasks. The experimental results show that our work surpasses the previous methods by a large margin in terms of success rate and execution efficiency. The experimental videos, code, and benchmark are released at https://github.com/MrKeee/COHERENT.
Authors: Jiachi Chen, Qingyuan Zhong, Yanlin Wang, Kaiwen Ning, Yongkun Liu, Zenan Xu, Zhe Zhao, Ting Chen, Zibin Zheng
Abstract: The emergence of Large Language Models (LLMs) has significantly influenced various aspects of software development activities. Despite their benefits, LLMs also pose notable risks, including the potential to generate harmful content and being abused by malicious developers to create malicious code. Several previous studies have focused on the ability of LLMs to resist the generation of harmful content that violates human ethical standards, such as biased or offensive content. However, there is no research evaluating the ability of LLMs to resist malicious code generation. To fill this gap, we propose RMCBench, the first benchmark comprising 473 prompts designed to assess the ability of LLMs to resist malicious code generation. This benchmark employs two scenarios: a text-to-code scenario, where LLMs are prompted with descriptions to generate code, and a code-to-code scenario, where LLMs translate or complete existing malicious code. Based on RMCBench, we conduct an empirical study on 11 representative LLMs to assess their ability to resist malicious code generation. Our findings indicate that current LLMs have a limited ability to resist malicious code generation with an average refusal rate of 40.36% in text-to-code scenario and 11.52% in code-to-code scenario. The average refusal rate of all LLMs in RMCBench is only 28.71%; ChatGPT-4 has a refusal rate of only 35.73%. We also analyze the factors that affect LLMs' ability to resist malicious code generation and provide implications for developers to enhance model robustness.
Authors: Bel\'en Serrano-Ant\'on, Mubashara Rehman, Niki Martinel, Michele Avanzo, Riccardo Spizzo, Giuseppe Fanetti, Alberto P. Mu\~nuzuri, Christian Micheloni
Abstract: For the planning of radiotherapy treatments for head and neck cancers, Computed Tomography (CT) scans of the patients are typically employed. However, in patients with head and neck cancer, the quality of standard CT scans generated using kilo-Voltage (kVCT) tube potentials is severely degraded by streak artifacts occurring in the presence of metallic implants such as dental fillings. Some radiotherapy devices offer the possibility of acquiring Mega-Voltage CT (MVCT) for daily patient setup verification, due to the higher energy of X-rays used, MVCT scans are almost entirely free from artifacts making them more suitable for radiotherapy treatment planning. In this study, we leverage the advantages of kVCT scans with those of MVCT scans (artifact-free). We propose a deep learning-based approach capable of generating artifact-free MVCT images from acquired kVCT images. The outcome offers the benefits of artifact-free MVCT images with enhanced soft tissue contrast, harnessing valuable information obtained through kVCT technology for precise therapy calibration. Our proposed method employs UNet-inspired model, and is compared with adversarial learning and transformer networks. This first and unique approach achieves remarkable success, with PSNR of 30.02 dB across the entire patient volume and 27.47 dB in artifact-affected regions exclusively. It is worth noting that the PSNR calculation excludes the background, concentrating solely on the region of interest.
Authors: Halid Abdulrahim Kadi, Jose Alex Chandy, Luis Figueredo, Kasim Terzi\'c, Praminda Caleb-Solly
Abstract: The fidelity gap between simulation-trained vision-based data-driven cloth neural controllers and real-world operation impedes reliable deployment of methods from simulation into physical trials. Real-world grasping errors, such as misgrasping and multilayer grasping, degrade their performance; additionally, some fabrics made of synthetic material also tend to stick to the commonly employed Franka Emika Panda's original gripper. Different approaches adopted various strategies to resolve these problems, further complicating real-world comparison between state-of-the-art methods. We propose DeepCloth-ROB$^2_{\text{QS}}$P&P with a simulation-to-reality transfer strategy Towel-Sim2Real and a cloth grasping protocol to consider and mitigate these grasping errors for robustly deploying quasi-static pick-and-place neural controllers in cloth shaping and demonstrate its generalisability across different deep-learning methods, fabric contexts and robot platforms. Our approach allows us to compare multiple neural controllers in a real environment for the first time, offering valuable insights to the cloth manipulation community.
Authors: Mrinal Verghese, Christopher Atkeson
Abstract: This study explores the utility of various internet data sources to select among a set of template robot behaviors to perform skills. Learning contact-rich skills involving tool use from internet data sources has typically been challenging due to the lack of physical information such as contact existence, location, areas, and force in this data. Prior works have generally used internet data and foundation models trained on this data to generate low-level robot behavior. We hypothesize that these data and models may be better suited to selecting among a set of basic robot behaviors to perform these contact-rich skills. We explore three methods of template selection: querying large language models, comparing video of robot execution to retrieved human video using features from a pretrained video encoder common in prior work, and performing the same comparison using features from an optic flow encoder trained on internet data. Our results show that LLMs are surprisingly capable template selectors despite their lack of visual information, optical flow encoding significantly outperforms video encoders trained with an order of magnitude more data, and important synergies exist between various forms of internet data for template selection. By exploiting these synergies, we create a template selector using multiple forms of internet data that achieves a 79\% success rate on a set of 16 different cooking skills involving tool-use.
Authors: Bingkun Yao, Ning Wang, Jie Zhou, Xi Wang, Hong Gao, Zhe Jiang, Nan Guan
Abstract: Bug localization in Verilog code is a crucial and time-consuming task during the verification of hardware design. Since introduction, Large Language Models (LLMs) have showed their strong programming capabilities. However, no work has yet considered using LLMs for bug localization in Verilog code. This paper presents Location-is-Key, an opensource LLM solution to locate functional errors in Verilog snippets. LiK achieves high localization accuracy, with a pass@1 localization accuracy of 93.3% on our test dataset based on RTLLM, surpassing GPT-4's 77.9% and comparable to Claude-3.5's 90.8%. Additionally, the bug location obtained by LiK significantly improves GPT-3.5's bug repair efficiency (Functional pass@1 increased from 40.39% to 58.92%), highlighting the importance of bug localization in LLM-based Verilog debugging. Compared to existing methods, LiK only requires the design specification and the erroneous code snippet, without the need for testbenches, assertions, or any other EDA tools. This research demonstrates the feasibility of using LLMs for Verilog error localization, thus providing a new direction for automatic Verilog code debugging.
Authors: Yuyan Chen, Yiwen Qian, Songzhou Yan, Jiyuan Jia, Zhixu Li, Yanghua Xiao, Xiaobo Li, Ming Yang, Qingpei Guo
Abstract: In the era of social media video platforms, popular ``hot-comments'' play a crucial role in attracting user impressions of short-form videos, making them vital for marketing and branding purpose. However, existing research predominantly focuses on generating descriptive comments or ``danmaku'' in English, offering immediate reactions to specific video moments. Addressing this gap, our study introduces \textsc{HotVCom}, the largest Chinese video hot-comment dataset, comprising 94k diverse videos and 137 million comments. We also present the \texttt{ComHeat} framework, which synergistically integrates visual, auditory, and textual data to generate influential hot-comments on the Chinese video dataset. Empirical evaluations highlight the effectiveness of our framework, demonstrating its excellence on both the newly constructed and existing datasets.
Authors: Mingqi Li, Karan Aggarwal, Yong Xie, Aitzaz Ahmad, Stephen Lau
Abstract: As LLMs evolve, significant effort is spent on manually crafting prompts. While existing prompt optimization methods automate this process, they rely solely on learning from incorrect samples, leading to a sub-optimal performance. Additionally, an unexplored challenge in the literature is prompts effective for prior models may not perform well on newer versions or different languages. We propose the Learning from Contrastive Prompts (LCP) framework to address these gaps, enhancing both prompt optimization and adaptation. LCP employs contrastive learning to generate effective prompts by analyzing patterns in good and bad prompt examples. Our evaluation on the Big-Bench Hard dataset shows that LCP has a win rate of over 76% over existing methods in prompt optimization and demonstrates strong adaptability across different model versions, families, and languages. LCP offers a systematic approach to prompt engineering, reducing manual effort in deploying LLMs across varied contexts.
Authors: Iwo Naglik, Mateusz Lango
Abstract: Aspect-Sentiment Triplet Extraction (ASTE) is a recently proposed task of aspect-based sentiment analysis that consists in extracting (aspect phrase, opinion phrase, sentiment polarity) triples from a given sentence. Recent state-of-the-art methods approach this task by first extracting all possible text spans from a given text, then filtering the potential aspect and opinion phrases with a classifier, and finally considering all their pairs with another classifier that additionally assigns sentiment polarity to them. Although several variations of the above scheme have been proposed, the common feature is that the final result is constructed by a sequence of independent classifier decisions. This hinders the exploitation of dependencies between extracted phrases and prevents the use of knowledge about the interrelationships between classifier predictions to improve performance. In this paper, we propose a new ASTE approach consisting of three transformer-inspired layers, which enables the modelling of dependencies both between phrases and between the final classifier decisions. Experimental results show that the method achieves higher performance in terms of F1 measure than other methods studied on popular benchmarks. In addition, we show that a simple pre-training technique further improves the performance of the model.
Authors: Rei Tamaru, Pei Li, Bin Ran
Abstract: Pedestrian trajectory prediction is essential for various applications in active traffic management, urban planning, traffic control, crowd management, and autonomous driving, aiming to enhance traffic safety and efficiency. Accurately predicting pedestrian trajectories requires a deep understanding of individual behaviors, social interactions, and road environments. Existing studies have developed various models to capture the influence of social interactions and road conditions on pedestrian trajectories. However, these approaches are limited by the lack of a comprehensive view of social interactions and road environments. To address these limitations and enhance the accuracy of pedestrian trajectory prediction, we propose a novel approach incorporating trip information as a new modality into pedestrian trajectory models. We propose RNTransformer, a generic model that utilizes crowd trip information to capture global information on social interactions. We incorporated RNTransformer with various socially aware local pedestrian trajectory prediction models to demonstrate its performance. Specifically, by leveraging a pre-trained RNTransformer when training different pedestrian trajectory prediction models, we observed improvements in performance metrics: a 1.3/2.2% enhancement in ADE/FDE on Social-LSTM, a 6.5/28.4% improvement on Social-STGCNN, and an 8.6/4.3% improvement on S-Implicit. Evaluation results demonstrate that RNTransformer significantly enhances the accuracy of various pedestrian trajectory prediction models across multiple datasets. Further investigation reveals that the RNTransformer effectively guides local models to more accurate directions due to the consideration of global information. By exploring crowd behavior within the road network, our approach shows great promise in improving pedestrian safety through accurate trajectory predictions.
Authors: Yixi Wu (Peter), Pengfei He (Peter), Zehao Wang (Peter), Shaowei Wang (Peter), Yuan Tian (Peter), Tse-Hsun (Peter), Chen
Abstract: Large language models (LLMs) like GitHub Copilot and ChatGPT have emerged as powerful tools for code generation, significantly enhancing productivity and accelerating software development. However, existing benchmarks primarily focus on general code generation without considering API-oriented code generation, i.e., generating code that invokes APIs from specific libraries. Given the growing demand for API-oriented code generation, there is a pressing need for a systematic and automated approach to evaluate LLM on API-oriented code generation. To address this gap, we propose AutoAPIEval, a lightweight and automated framework designed to evaluate the capabilities of LLMs in API-oriented code generation. Our framework works with any library that provides API documentation and focuses on two unit tasks: API recommendation and code example generation, along with four metrics to evaluate the generated APIs and code examples, such as the proportion of incorrect API recommendations for Task 1, and the proportion of code examples where no specific API is invoked and uncompilable/unexecutable code examples for Task 2. In addition, we conducted a case study on three LLMs (ChatGPT, MagiCoder, and DeepSeek Coder) and Java Runtime Environment 8 to demonstrate the framework's effectiveness. Our findings reveal substantial variability in LLM performance across tasks, with ChatGPT adhering better to instructions, while sharing similar effectiveness in code example generation with its counterparts (i.e., MagiCoder and DeekSeek Coder). We also identify key factors associated with code quality, such as API popularity and model confidence, and build classifiers that achieve high accuracy in detecting incorrect API recommendations and erroneous code examples. Retrieval-augmented generation enhances the quality of code generated by LLMs, though its effectiveness varies across different LLMs.
Authors: Junqing He, Liang Zhu, Qi Wei, Rui Wang, Jiaxing Zhang
Abstract: Long-term memory is so important for chatbots and dialogue systems (DS) that researchers have developed numerous memory-augmented DS. However, their evaluation methods are different from the real situation in human conversation. They only measured the accuracy of factual information or the perplexity of generated responses given a query, which hardly reflected their performance. Moreover, they only consider passive memory retrieval based on similarity, neglecting diverse memory-recalling paradigms in humans, e.g. emotions and surroundings. To bridge the gap, we construct a novel benchmark covering various memory recalling paradigms based on cognitive science and psychology theory. The Memory Benchmark (MemBench) contains two tasks according to the two-phrase theory in cognitive science: memory retrieval, memory recognition and injection. The benchmark considers both passive and proactive memory recalling based on meta information for the first time. In addition, novel scoring aspects are proposed to comprehensively measure the generated responses. Results from the strongest embedding models and LLMs on MemBench show that there is plenty of room for improvement in existing dialogue systems. Extensive experiments also reveal the correlation between memory injection and emotion supporting (ES) skillfulness, and intimacy. Our code and dataset will be released.
Authors: Guanhua Wang, Chengming Zhang, Zheyu Shen, Ang Li, Olatunji Ruwase
Abstract: Given the popularity of generative AI, Large Language Models (LLMs) often consume hundreds or thousands of GPUs for parallelizing and accelerating the training process. Communication overhead becomes more pronounced when training LLMs at scale. To eliminate communication overhead in distributed LLM training, we propose Domino, which provides a generic scheme to hide communication behind computation. By breaking data dependency of a single batch training into smaller independent pieces, Domino pipelines these independent pieces training and provides generic strategy of fine-grained communication and computation overlapping. Extensive results show that, comparing with Megatron-LM, Domino achieves up to 1.3x speedup for LLM training on Nvidia DGX-H100 GPUs.
Authors: Jon Saad-Falcon, Adrian Gamarra Lafuente, Shlok Natarajan, Nahum Maru, Hristo Todorov, E. Kelly Buchanan, Mayee Chen, Neel Guha, Christopher R\'e, Azalia Mirhoseini
Abstract: Inference-time techniques are emerging as highly effective tools to increase large language model (LLM) capabilities. However, there is still limited understanding of the best practices for developing systems that combine inference-time techniques with one or more LLMs, with challenges including: (1) effectively allocating inference compute budget, (2) understanding the interactions between different combinations of inference-time techniques and their impact on downstream performance, and 3) efficiently searching over the large space of model choices, inference-time techniques, and their compositions. To address these challenges, we introduce Archon, an automated framework for designing inference-time architectures. Archon defines an extensible design space, encompassing methods such as generation ensembling, multi-sampling, ranking, fusion, critiquing, verification, and unit testing. It then transforms the problem of selecting and combining LLMs and inference-time techniques into a hyperparameter optimization objective. To optimize this objective, we introduce automated Inference-Time Architecture Search (ITAS) algorithms. Given target benchmark(s), an inference compute budget, and available LLMs, ITAS outputs optimized architectures. We evaluate Archon architectures across a wide range of instruction-following and reasoning benchmarks, including MT-Bench, Arena-Hard-Auto, AlpacaEval 2.0, MixEval, MixEval Hard, MATH, and CodeContests. We show that automatically designed inference-time architectures by Archon outperform strong models such as GPT-4o and Claude 3.5 Sonnet on these benchmarks, achieving an average increase of 14.1 and 10.3 percentage points with all-source models and open-source models, respectively. We make our code and datasets available publicly on Github: https://github.com/ScalingIntelligence/Archon.
Authors: Yuhang Xiao, Yudi Lin, Ming-Chang Chiu
Abstract: Large Vision-Language Models (LVLMs) evolve rapidly as Large Language Models (LLMs) was equipped with vision modules to create more human-like models. However, we should carefully evaluate their applications in different domains, as they may possess undesired biases. Our work studies the potential behavioral biases of LVLMs from a behavioral finance perspective, an interdisciplinary subject that jointly considers finance and psychology. We propose an end-to-end framework, from data collection to new evaluation metrics, to assess LVLMs' reasoning capabilities and the dynamic behaviors manifested in two established human financial behavioral biases: recency bias and authority bias. Our evaluations find that recent open-source LVLMs such as LLaVA-NeXT, MobileVLM-V2, Mini-Gemini, MiniCPM-Llama3-V 2.5 and Phi-3-vision-128k suffer significantly from these two biases, while the proprietary model GPT-4o is negligibly impacted. Our observations highlight directions in which open-source models can improve. The code is available at https://github.com/mydcxiao/vlm_behavioral_fin.
Authors: Yuanhang Li, Qi Mao, Lan Chen, Zhen Fang, Lei Tian, Xinyan Xiao, Libiao Jin, Hua Wu
Abstract: Recent advancements in text-to-video (T2V) generation using diffusion models have garnered significant attention. However, existing T2V models primarily focus on simple scenes featuring a single object performing a single motion. Challenges arise in scenarios involving multiple objects with distinct motions, often leading to incorrect video-text alignment between subjects and their corresponding motions. To address this challenge, we propose \textbf{S$^2$AG-Vid}, a training-free inference-stage optimization method that improves the alignment of multiple objects with their corresponding motions in T2V models. S$^2$AG-Vid initially applies a spatial position-based, cross-attention (CA) constraint in the early stages of the denoising process, facilitating multiple nouns distinctly attending to the correct subject regions. To enhance the motion-subject binding, we implement a syntax-guided contrastive constraint in the subsequent denoising phase, aimed at improving the correlations between the CA maps of verbs and their corresponding nouns.Both qualitative and quantitative evaluations demonstrate that the proposed framework significantly outperforms baseline approaches, producing higher-quality videos with improved subject-motion consistency.
Authors: Bryce T. Bolin, Michael W. Coughlin
Abstract: In this chapter, we will discuss the use of Machine Learning methods for the identification and localization of cometary activity for Solar System objects in ground and in space-based wide-field all-sky surveys. We will begin the chapter by discussing the challenges of identifying known and unknown active, extended Solar System objects in the presence of stellar-type sources and the application of classical pre-ML identification techniques and their limitations. We will then transition to the discussion of implementing ML techniques to address the challenge of extended object identification. We will finish with prospective future methods and the application to future surveys such as the Vera C. Rubin Observatory.
Authors: B. T. Bolin, F. J. Masci, M. W. Coughlin, D. A. Duev, \v{Z}. Ivezi\'c, R. L. Jones, P. Yoachim, T. Ahumada, V. Bhalerao, H. Choudhary, C. Contreras, Y. -C. Cheng, C. M. Copperwheat, K. Deshmukh, C. Fremling, M. Granvik, K. K. Hardegree-Ullman, A. Y. Q. Ho, R. Jedicke, M. Kasliwal, H. Kumar, Z. -Y. Lin, A. Mahabal, A. Monson, J. D. Neill, D. Nesvorn\'y, D. A. Perley, J. N. Purdum, R. Quimby, E. Serabyn, K. Sharma, V. Swain
Abstract: Near-sun sky twilight observations allow for the detection of asteroid interior to the orbit of Venus (Aylos), the Earth (Atiras), and comets. We present the results of observations with the Palomar 48-inch telescope (P48)/Zwicky Transient Facility (ZTF) camera in 30 s r-band exposures taken during evening astronomical twilight from 2019 Sep 20 to 2022 March 7 and during morning astronomical twilight sky from 2019 Sep 21 to 2022 Sep 29. More than 46,000 exposures were taken in evening and morning astronomical twilight within 31 to 66 degrees from the Sun with an r-band limiting magnitude between 18.1 and 20.9. The twilight pointings show a slight seasonal dependence in limiting magnitude and ability to point closer towards the Sun, with limiting magnitude slightly improving during summer. In total, the one Aylo, (594913) 'Ayl\'o'chaxnim, and 4 Atiras, 2020 OV1, 2021 BS1, 2021 PB2, and 2021 VR3, were discovered in evening and morning twilight observations. Additional twilight survey discoveries also include 6 long-period comets: C/2020 T2, C/2020 V2, C/2021 D2, C/2021 E3, C/2022 E3, and C/2022 P3, and two short-period comets: P/2021 N1 and P/2022 P2 using deep learning comet detection pipelines. The P48/ZTF twilight survey also recovered 11 known Atiras, one Aylo, three short-period comes, two long-period comets, and one interstellar object. Lastly, the Vera Rubin Observatory will conduct a twilight survey starting in its first year of operations and will cover the sky within 45 degrees of the Sun. Twilight surveys such as those by ZTF and future surveys will provide opportunities for discovering asteroids inside the orbits of Earth and Venus.
Authors: Benjamin Feuer, Micah Goldblum, Teresa Datta, Sanjana Nambiar, Raz Besaleli, Samuel Dooley, Max Cembalest, John P. Dickerson
Abstract: The release of ChatGPT in November 2022 sparked an explosion of interest in post-training and an avalanche of new preference optimization (PO) methods. These methods claim superior alignment by virtue of better correspondence with human pairwise preferences, often measured by LLM judges. In this work, we attempt to answer the following question -- do LLM-judge preferences translate to progress on other, more concrete metrics for alignment, and if not, why not? We define a concrete metric for alignment, and introduce SOS-Bench, the largest standardized, reproducible LLM meta-benchmark to date. We find that (1) LLM-judgments do not correlate with concrete measures of safety, world knowledge, and instruction following; (2) LLM judges have powerful implicit biases, prioritizing style over factuality and safety; and (3) the supervised fine-tuning (SFT) stage of post-training, and not the PO stage, has the greatest impact on alignment, with data scaling and prompt diversity as the driving factors. Our codebase and complete results can be found at https://github.com/penfever/sos-bench.
Authors: Yizhi Li, Ge Zhang, Yinghao Ma, Ruibin Yuan, Kang Zhu, Hangyu Guo, Yiming Liang, Jiaheng Liu, Jian Yang, Siwei Wu, Xingwei Qu, Jinjie Shi, Xinyue Zhang, Zhenzhu Yang, Xiangzhou Wang, Zhaoxiang Zhang, Zachary Liu, Emmanouil Benetos, Wenhao Huang, Chenghua Lin
Abstract: Recent advancements in multimodal large language models (MLLMs) have aimed to integrate and interpret data across diverse modalities. However, the capacity of these models to concurrently process and reason about multiple modalities remains inadequately explored, partly due to the lack of comprehensive modality-wise benchmarks. We introduce OmniBench, a novel benchmark designed to rigorously evaluate models' ability to recognize, interpret, and reason across visual, acoustic, and textual inputs simultaneously. We define models capable of such tri-modal processing as omni-language models (OLMs). OmniBench is distinguished by high-quality human annotations, ensuring that accurate responses require integrated understanding and reasoning across all three modalities. Our main findings reveal that: i) open-source OLMs exhibit critical limitations in instruction-following and reasoning capabilities within tri-modal contexts; and ii) the baseline models perform poorly (below 50% accuracy) even when provided with alternative textual representations of images and audio. These results suggest that the ability to construct a consistent context from text, image, and audio is often overlooked in existing MLLM training paradigms. We advocate for future research to focus on developing more robust tri-modal integration techniques and training strategies to enhance OLM performance across diverse modalities. The codes and live leaderboard could be found at https://m-a-p.ai/OmniBench.
Authors: Yunfei Xie, Juncheng Wu, Haoqin Tu, Siwei Yang, Bingchen Zhao, Yongshuo Zong, Qiao Jin, Cihang Xie, Yuyin Zhou
Abstract: Large language models (LLMs) have exhibited remarkable capabilities across various domains and tasks, pushing the boundaries of our knowledge in learning and cognition. The latest model, OpenAI's o1, stands out as the first LLM with an internalized chain-of-thought technique using reinforcement learning strategies. While it has demonstrated surprisingly strong capabilities on various general language tasks, its performance in specialized fields such as medicine remains unknown. To this end, this report provides a comprehensive exploration of o1 on different medical scenarios, examining 3 key aspects: understanding, reasoning, and multilinguality. Specifically, our evaluation encompasses 6 tasks using data from 37 medical datasets, including two newly constructed and more challenging question-answering (QA) tasks based on professional medical quizzes from the New England Journal of Medicine (NEJM) and The Lancet. These datasets offer greater clinical relevance compared to standard medical QA benchmarks such as MedQA, translating more effectively into real-world clinical utility. Our analysis of o1 suggests that the enhanced reasoning ability of LLMs may (significantly) benefit their capability to understand various medical instructions and reason through complex clinical scenarios. Notably, o1 surpasses the previous GPT-4 in accuracy by an average of 6.2% and 6.6% across 19 datasets and two newly created complex QA scenarios. But meanwhile, we identify several weaknesses in both the model capability and the existing evaluation protocols, including hallucination, inconsistent multilingual ability, and discrepant metrics for evaluation. We release our raw data and model outputs at https://ucsc-vlaa.github.io/o1_medicine/ for future research.
Authors: Shaopeng Wei, Jun Wang, Yu Zhao, Xingyan Chen, Qing Li, Fuzhen Zhuang, Ji Liu, Fuji Ren, Gang Kou
Abstract: Graph learning is a prevalent domain that endeavors to learn the intricate relationships among nodes and the topological structure of graphs. Over the years, graph learning has transcended from graph theory to graph data mining. With the advent of representation learning, it has attained remarkable performance in diverse scenarios. Owing to its extensive application prospects, graph learning attracts copious attention. While some researchers have accomplished impressive surveys on graph learning, they failed to connect related objectives, methods, and applications in a more coherent way. As a result, they did not encompass current ample scenarios and challenging problems due to the rapid expansion of graph learning. Particularly, large language models have recently had a disruptive effect on human life, but they also show relative weakness in structured scenarios. The question of how to make these models more powerful with graph learning remains open. Our survey focuses on the most recent advancements in integrating graph learning with pre-trained language models, specifically emphasizing their application within the domain of large language models. Different from previous surveys on graph learning, we provide a holistic review that analyzes current works from the perspective of graph structure, and discusses the latest applications, trends, and challenges in graph learning. Specifically, we commence by proposing a taxonomy and then summarize the methods employed in graph learning. We then provide a detailed elucidation of mainstream applications. Finally, we propose future directions.
Authors: Surya Narayanan Hari, Rex Liu, Matt Thomson
Abstract: Currently, over a thousand LLMs exist that are multi-purpose and are capable of performing real world tasks, including Q&A, text summarization, content generation, etc. However, accessibility, scale and reliability of free models prevents them from being widely deployed in everyday use cases. To address the first two issues of access and scale, organisations such as HuggingFace have created model repositories where users have uploaded model weights and quantized versions of models trained using different paradigms, as well as model cards describing their training process. While some models report performance on commonly used benchmarks, not all do, and interpreting the real world impact of trading off performance on a benchmark for model deployment cost, is unclear. Here, we show that a herd of open source models can match or exceed the performance of proprietary models via an intelligent router. We show that a Herd of open source models is able to match the accuracy of ChatGPT, despite being composed of models that are effectively 2.5x smaller. We show that in cases where GPT is not able to answer the query, Herd is able to identify a model that can, at least 40% of the time.
Authors: Daniele Regoli, Alessandro Castelnovo, Nicole Inverardi, Gabriele Nanino, Ilaria Penco
Abstract: In recent years, the increase in the usage and efficiency of Artificial Intelligence and, more in general, of Automated Decision-Making systems has brought with it an increasing and welcome awareness of the risks associated with such systems. One of such risks is that of perpetuating or even amplifying bias and unjust disparities present in the data from which many of these systems learn to adjust and optimise their decisions. This awareness has on the one hand encouraged several scientific communities to come up with more and more appropriate ways and methods to assess, quantify, and possibly mitigate such biases and disparities. On the other hand, it has prompted more and more layers of society, including policy makers, to call for fair algorithms. We believe that while many excellent and multidisciplinary research is currently being conducted, what is still fundamentally missing is the awareness that having fair algorithms is per se a nearly meaningless requirement that needs to be complemented with many additional social choices to become actionable. Namely, there is a hiatus between what the society is demanding from Automated Decision-Making systems, and what this demand actually means in real-world scenarios. In this work, we outline the key features of such a hiatus and pinpoint a set of crucial open points that we as a society must address in order to give a concrete meaning to the increasing demand of fairness in Automated Decision-Making systems.
Authors: Jiri Nemecek, Tomas Pevny, Jakub Marecek
Abstract: Explainability of decisions made by AI systems is driven by both recent regulation and user demand. These decisions are often explainable only \emph{post hoc}, after the fact. In counterfactual explanations, one may ask what constitutes the best counterfactual explanation. Clearly, multiple criteria must be taken into account, although "distance from the sample" is a key criterion. Recent methods that consider the plausibility of a counterfactual seem to sacrifice this original objective. Here, we present a system that provides high-likelihood explanations that are, at the same time, close and sparse. We show that the search for the most likely explanations satisfying many common desiderata for counterfactual explanations can be modeled using mixed-integer optimization (MIO). In the process, we propose an MIO formulation of a Sum-Product Network (SPN) and use the SPN to estimate the likelihood of a counterfactual, which can be of independent interest.
Authors: Louis Mozart Kamdem Teyou, Caglar Demir, Axel-Cyrille Ngonga Ngomo
Abstract: Clifford algebras are a natural generalization of the real numbers, the complex numbers, and the quaternions. So far, solely Clifford algebras of the form $Cl_{p,q}$ (i.e., algebras without nilpotent base vectors) have been studied in the context of knowledge graph embeddings. We propose to consider nilpotent base vectors with a nilpotency index of two. In these spaces, denoted $Cl_{p,q,r}$, allows generalizing over approaches based on dual numbers (which cannot be modelled using $Cl_{p,q}$) and capturing patterns that emanate from the absence of higher-order interactions between real and complex parts of entity embeddings. We design two new models for the discovery of the parameters $p$, $q$, and $r$. The first model uses a greedy search to optimize $p$, $q$, and $r$. The second predicts $(p, q,r)$ based on an embedding of the input knowledge graph computed using neural networks. The results of our evaluation on seven benchmark datasets suggest that nilpotent vectors can help capture embeddings better. Our comparison against the state of the art suggests that our approach generalizes better than other approaches on all datasets w.r.t. the MRR it achieves on validation data. We also show that a greedy search suffices to discover values of $p$, $q$ and $r$ that are close to optimal.
Authors: Hao Tang, Darren Key, Kevin Ellis
Abstract: We give a model-based agent that builds a Python program representing its knowledge of the world based on its interactions with the environment. The world model tries to explain its interactions, while also being optimistic about what reward it can achieve. We define this optimism as a logical constraint between a program and a planner. We study our agent on gridworlds, and on task planning, finding our approach is more sample-efficient compared to deep RL, more compute-efficient compared to ReAct-style agents, and that it can transfer its knowledge across environments by editing its code.
Authors: Adrita Barua, Cara Widmer, Pascal Hitzler
Abstract: Explainable Artificial Intelligence (XAI) poses a significant challenge in providing transparent and understandable insights into complex AI models. Traditional post-hoc algorithms, while useful, often struggle to deliver interpretable explanations. Concept-based models offer a promising avenue by incorporating explicit representations of concepts to enhance interpretability. However, existing research on automatic concept discovery methods is often limited by lower-level concepts, costly human annotation requirements, and a restricted domain of background knowledge. In this study, we explore the potential of a Large Language Model (LLM), specifically GPT-4, by leveraging its domain knowledge and common-sense capability to generate high-level concepts that are meaningful as explanations for humans, for a specific setting of image classification. We use minimal textual object information available in the data via prompting to facilitate this process. To evaluate the output, we compare the concepts generated by the LLM with two other methods: concepts generated by humans and the ECII heuristic concept induction system. Since there is no established metric to determine the human understandability of concepts, we conducted a human study to assess the effectiveness of the LLM-generated concepts. Our findings indicate that while human-generated explanations remain superior, concepts derived from GPT-4 are more comprehensible to humans compared to those generated by ECII.
Authors: Ritesh Chandra, Shashi Shekhar Kumar, Rushil Patra, Sonali Agarwal
Abstract: Forests are crucial for ecological balance, but wildfires, a major cause of forest loss, pose significant risks. Fire weather indices, which assess wildfire risk and predict resource demands, are vital. With the rise of sensor networks in fields like healthcare and environmental monitoring, semantic sensor networks are increasingly used to gather climatic data such as wind speed, temperature, and humidity. However, processing these data streams to determine fire weather indices presents challenges, underscoring the growing importance of effective forest fire detection. This paper discusses using Apache Spark for early forest fire detection, enhancing fire risk prediction with meteorological and geographical data. Building on our previous development of Semantic Sensor Network (SSN) ontologies and Semantic Web Rules Language (SWRL) for managing forest fires in Monesterial Natural Park, we expanded SWRL to improve a Decision Support System (DSS) using a Large Language Models (LLMs) and Spark framework. We implemented real-time alerts with Spark streaming, tailored to various fire scenarios, and validated our approach using ontology metrics, query-based evaluations, LLMs score precision, F1 score, and recall measures.
Authors: Seyed Arash Sheikholeslam, Andre Ivanov
Abstract: In this paper, we introduce SynthAI, a new method for the automated creation of High-Level Synthesis (HLS) designs. SynthAI integrates ReAct agents, Chain-of-Thought (CoT) prompting, web search technologies, and the Retrieval-Augmented Generation (RAG) framework within a structured decision graph. This innovative approach enables the systematic decomposition of complex hardware design tasks into multiple stages and smaller, manageable modules. As a result, SynthAI produces synthesizable designs that closely adhere to user-specified design objectives and functional requirements. We further validate the capabilities of SynthAI through several case studies, highlighting its proficiency in generating complex, multi-module logic designs from a single initial prompt. The SynthAI code is provided via the following repo: \url{https://github.com/sarashs/FPGA_AGI}
Authors: Liang Zhang, Jionghao Lin, Ziyi Kuang, Sheng Xu, Xiangen Hu
Abstract: Dialogue-based Intelligent Tutoring Systems (ITSs) have significantly advanced adaptive and personalized learning by automating sophisticated human tutoring strategies within interactive dialogues. However, replicating the nuanced patterns of expert human communication remains a challenge in Natural Language Processing (NLP). Recent advancements in NLP, particularly Large Language Models (LLMs) such as OpenAI's GPT-4, offer promising solutions by providing human-like and context-aware responses based on extensive pre-trained knowledge. Motivated by the effectiveness of LLMs in various educational tasks (e.g., content creation and summarization, problem-solving, and automated feedback provision), our study introduces the Socratic Playground for Learning (SPL), a dialogue-based ITS powered by the GPT-4 model, which employs the Socratic teaching method to foster critical thinking among learners. Through extensive prompt engineering, SPL can generate specific learning scenarios and facilitates efficient multi-turn tutoring dialogues. The SPL system aims to enhance personalized and adaptive learning experiences tailored to individual needs, specifically focusing on improving critical thinking skills. Our pilot experimental results from essay writing tasks demonstrate SPL has the potential to improve tutoring interactions and further enhance dialogue-based ITS functionalities. Our study, exemplified by SPL, demonstrates how LLMs enhance dialogue-based ITSs and expand the accessibility and efficacy of educational technologies.
Authors: Rujia Shen, Yaoxion Lin, Liangliang Liu, Boran Wang, Yi Guan, Yang Yang, Jingchi Jiang
Abstract: Time series forecasting (TSF) plays a crucial role in various applications, including electricity transformation, medical monitoring, and crop growth. Despite the advancements in deep learning methods for TSF, their capacity to predict long-term series remains constrained. This limitation arises from the failure to account for both intra- and inter-variable variations meanwhile. To mitigate this challenge, we introduce the Fi$^2$VBlock, which leverages a \textbf{F}requency domain perspective to capture \textbf{i}ntra- and \textbf{i}nter-variable \textbf{V}ariations. After transforming into the frequency domain via the Frequency Transform Module, the Frequency Cross Attention between the real and imaginary parts is designed to obtain enhanced frequency representations and capture intra-variable variations. Furthermore, Inception blocks are employed to integrate information, thus capturing correlations across different variables. Our backbone network, Fi$^2$VTS, employs a residual architecture by concatenating multiple Fi$^2$VBlocks, thereby preventing degradation issues. Theoretically, we demonstrate that Fi$^2$VTS achieves a substantial reduction in both time and memory complexity, decreasing from $\mathcal{O}(L^2)$ to $\mathcal{O}(L)$ per Fi$^2$VBlock computation. Empirical evaluations on three benchmark datasets reveal that Fi$^2$VTS delivers an overall relative Mean Squared Error (MSE) reduction of 30\% and an overall relative Mean Absolute Error (MAE) reduction of 22\% when compared to the latest state-of-the-art methods. The implementation code is accessible at \url{https://github.com/HITshenrj/Fi2VTS}.
Authors: Zhiyuan Li, Yanfeng Lu, Yao Mu, Hong Qiao
Abstract: Vision Language Navigation in Continuous Environments (VLN-CE) represents a frontier in embodied AI, demanding agents to navigate freely in unbounded 3D spaces solely guided by natural language instructions. This task introduces distinct challenges in multimodal comprehension, spatial reasoning, and decision-making. To address these challenges, we introduce Cog-GA, a generative agent founded on large language models (LLMs) tailored for VLN-CE tasks. Cog-GA employs a dual-pronged strategy to emulate human-like cognitive processes. Firstly, it constructs a cognitive map, integrating temporal, spatial, and semantic elements, thereby facilitating the development of spatial memory within LLMs. Secondly, Cog-GA employs a predictive mechanism for waypoints, strategically optimizing the exploration trajectory to maximize navigational efficiency. Each waypoint is accompanied by a dual-channel scene description, categorizing environmental cues into 'what' and 'where' streams as the brain. This segregation enhances the agent's attentional focus, enabling it to discern pertinent spatial information for navigation. A reflective mechanism complements these strategies by capturing feedback from prior navigation experiences, facilitating continual learning and adaptive replanning. Extensive evaluations conducted on VLN-CE benchmarks validate Cog-GA's state-of-the-art performance and ability to simulate human-like navigation behaviors. This research significantly contributes to the development of strategic and interpretable VLN-CE agents.
Authors: Zhiyuan Li, Yanfeng Lv, Ziqin Tu, Di Shang, Hong Qiao
Abstract: Vision-language navigation (VLN) is a critical domain within embedded intelligence, requiring agents to navigate 3D environments based on natural language instructions. Traditional VLN research has focused on improving environmental understanding and decision accuracy. However, these approaches often exhibit a significant performance gap when agents are deployed in novel environments, mainly due to the limited diversity of training data. Expanding datasets to cover a broader range of environments is impractical and costly. We propose the Vision-Language Navigation with Continual Learning (VLNCL) paradigm to address this challenge. In this paradigm, agents incrementally learn new environments while retaining previously acquired knowledge. VLNCL enables agents to maintain an environmental memory and extract relevant knowledge, allowing rapid adaptation to new environments while preserving existing information. We introduce a novel dual-loop scenario replay method (Dual-SR) inspired by brain memory replay mechanisms integrated with VLN agents. This method facilitates consolidating past experiences and enhances generalization across new tasks. By utilizing a multi-scenario memory buffer, the agent efficiently organizes and replays task memories, thereby bolstering its ability to adapt quickly to new environments and mitigating catastrophic forgetting. Our work pioneers continual learning in VLN agents, introducing a novel experimental setup and evaluation metrics. We demonstrate the effectiveness of our approach through extensive evaluations and establish a benchmark for the VLNCL paradigm. Comparative experiments with existing continual learning and VLN methods show significant improvements, achieving state-of-the-art performance in continual learning ability and highlighting the potential of our approach in enabling rapid adaptation while preserving prior knowledge.
Authors: Peng Chen, Pi Bu, Jun Song, Yuan Gao, Bo Zheng
Abstract: Recently, large language model (LLM)-based agents have made significant advances across various fields. One of the most popular research areas involves applying these agents to video games. Traditionally, these methods have relied on game APIs to access in-game environmental and action data. However, this approach is limited by the availability of APIs and does not reflect how humans play games. With the advent of vision language models (VLMs), agents now have enhanced visual understanding capabilities, enabling them to interact with games using only visual inputs. Despite these advances, current approaches still face challenges in action-oriented tasks, particularly in action role-playing games (ARPGs), where reinforcement learning methods are prevalent but suffer from poor generalization and require extensive training. To address these limitations, we select an ARPG, ``Black Myth: Wukong'', as a research platform to explore the capability boundaries of existing VLMs in scenarios requiring visual-only input and complex action output. We define 12 tasks within the game, with 75% focusing on combat, and incorporate several state-of-the-art VLMs into this benchmark. Additionally, we will release a human operation dataset containing recorded gameplay videos and operation logs, including mouse and keyboard actions. Moreover, we propose a novel VARP (Vision Action Role-Playing) agent framework, consisting of an action planning system and a visual trajectory system. Our framework demonstrates the ability to perform basic tasks and succeed in 90% of easy and medium-level combat scenarios. This research aims to provide new insights and directions for applying multimodal agents in complex action game environments. The code and datasets will be made available at https://varp-agent.github.io/.
Authors: Daniel N. Nissani (Nissensohn)
Abstract: After four decades of research there still exists a Classification accuracy gap of about 20% between our best Unsupervisedly Learned Representations methods and the accuracy rates achieved by intelligent animals. It thus may well be that we are looking in the wrong direction. A possible solution to this puzzle is presented. We demonstrate that Reinforcement Learning can learn representations which achieve the same accuracy as that of animals. Our main modest contribution lies in the observations that: a. when applied to a real world environment Reinforcement Learning does not require labels, and thus may be legitimately considered as Unsupervised Learning, and b. in contrast, when Reinforcement Learning is applied in a simulated environment it does inherently require labels and should thus be generally be considered as Supervised Learning. The corollary of these observations is that further search for Unsupervised Learning competitive paradigms which may be trained in simulated environments may be futile.
Authors: Wenhao Li, Mengyuan Liu, Hong Liu, Tianyu Guo, Ti Wang, Hao Tang, Nicu Sebe
Abstract: Modern multi-layer perceptron (MLP) models have shown competitive results in learning visual representations without self-attention. However, existing MLP models are not good at capturing local details and lack prior knowledge of human body configurations, which limits their modeling power for skeletal representation learning. To address these issues, we propose a simple yet effective graph-reinforced MLP-Like architecture, named GraphMLP, that combines MLPs and graph convolutional networks (GCNs) in a global-local-graphical unified architecture for 3D human pose estimation. GraphMLP incorporates the graph structure of human bodies into an MLP model to meet the domain-specific demand of the 3D human pose, while allowing for both local and global spatial interactions. Furthermore, we propose to flexibly and efficiently extend the GraphMLP to the video domain and show that complex temporal dynamics can be effectively modeled in a simple way with negligible computational cost gains in the sequence length. To the best of our knowledge, this is the first MLP-Like architecture for 3D human pose estimation in a single frame and a video sequence. Extensive experiments show that the proposed GraphMLP achieves state-of-the-art performance on two datasets, i.e., Human3.6M and MPI-INF-3DHP. Code and models are available at https://github.com/Vegetebird/GraphMLP.
Authors: Lin Shi, Qiyuan Liu, Bei Peng
Abstract: Relative overgeneralization (RO) is a pathology that can arise in cooperative multi-agent tasks when the optimal joint action's utility falls below that of a sub-optimal joint action. RO can cause the agents to get stuck into local optima or fail to solve cooperative tasks requiring significant coordination between agents within a given timestep. In this work, we empirically find that, in multi-agent reinforcement learning (MARL), both value-based and policy gradient MARL algorithms can suffer from RO and fail to learn effective coordination policies. To better overcome RO, we propose a novel approach called curriculum learning for relative overgeneralization (CURO). To solve a target task that exhibits strong RO, in CURO, we first fine-tune the reward function of the target task to generate source tasks to train the agent. Then, to effectively transfer the knowledge acquired in one task to the next, we use a transfer learning method that combines value function transfer with buffer transfer, which enables more efficient exploration in the target task. CURO is general and can be applied to both value-based and policy gradient MARL methods. We demonstrate that, when applied to QMIX, HAPPO, and HATRPO, CURO can successfully overcome severe RO, achieve improved performance, and outperform baseline methods in a variety of challenging cooperative multi-agent tasks.
Authors: Zhipeng Liang, Xiaoteng Ma, Jose Blanchet, Jiheng Zhang, Zhengyuan Zhou
Abstract: To mitigate the limitation that the classical reinforcement learning (RL) framework heavily relies on identical training and test environments, Distributionally Robust RL (DRRL) has been proposed to enhance performance across a range of environments, possibly including unknown test environments. As a price for robustness gain, DRRL involves optimizing over a set of distributions, which is inherently more challenging than optimizing over a fixed distribution in the non-robust case. Existing DRRL algorithms are either model-based or fail to learn from a single sample trajectory. In this paper, we design a first fully model-free DRRL algorithm, called distributionally robust Q-learning with single trajectory (DRQ). We delicately design a multi-timescale framework to fully utilize each incrementally arriving sample and directly learn the optimal distributionally robust policy without modelling the environment, thus the algorithm can be trained along a single trajectory in a model-free fashion. Despite the algorithm's complexity, we provide asymptotic convergence guarantees by generalizing classical stochastic approximation tools. Comprehensive experimental results demonstrate the superior robustness and sample complexity of our proposed algorithm, compared to non-robust methods and other robust RL algorithms.
Authors: Carmel Fiscko, Soummya Kar, Bruno Sinopoli
Abstract: This work studies a multi-agent Markov decision process (MDP) that can undergo agent dropout and the computation of policies for the post-dropout system based on control and sampling of the pre-dropout system. The central planner's objective is to find an optimal policy that maximizes the value of the expected system given a priori knowledge of the agents' dropout probabilities. For MDPs with a certain transition independence and reward separability structure, we assume that removing agents from the system forms a new MDP comprised of the remaining agents with new state and action spaces, transition dynamics that marginalize the removed agents, and rewards that are independent of the removed agents. We first show that under these assumptions, the value of the expected post-dropout system can be represented by a single MDP; this "robust MDP" eliminates the need to evaluate all $2^N$ realizations of the system, where N denotes the number of agents. More significantly, in a model-free context, it is shown that the robust MDP value can be estimated with samples generated by the pre-dropout system, meaning that robust policies can be found before dropout occurs. This fact is used to propose a policy importance sampling (IS) routine that performs policy evaluation for dropout scenarios while controlling the existing system with good pre-dropout policies. The policy IS routine produces value estimates for both the robust MDP and specific post-dropout system realizations and is justified with exponential confidence bounds. Finally, the utility of this approach is verified in simulation, showing how structural properties of agent dropout can help a controller find good post-dropout policies before dropout occurs.
Authors: Aditya Prakash, Matthew Chang, Matthew Jin, Ruisen Tu, Saurabh Gupta
Abstract: Prior works for reconstructing hand-held objects from a single image train models on images paired with 3D shapes. Such data is challenging to gather in the real world at scale. Consequently, these approaches do not generalize well when presented with novel objects in in-the-wild settings. While 3D supervision is a major bottleneck, there is an abundance of a) in-the-wild raw video data showing hand-object interactions and b) synthetic 3D shape collections. In this paper, we propose modules to leverage 3D supervision from these sources to scale up the learning of models for reconstructing hand-held objects. Specifically, we extract multiview 2D mask supervision from videos and 3D shape priors from shape collections. We use these indirect 3D cues to train occupancy networks that predict the 3D shape of objects from a single RGB image. Our experiments in the challenging object generalization setting on in-the-wild MOW dataset show 11.6% relative improvement over models trained with 3D supervision on existing datasets.
Authors: Qingyu Chen, Jingcheng Du, Yan Hu, Vipina Kuttichi Keloth, Xueqing Peng, Kalpana Raja, Rui Zhang, Zhiyong Lu, Hua Xu
Abstract: Biomedical literature is growing rapidly, making it challenging to curate and extract knowledge manually. Biomedical natural language processing (BioNLP) techniques that can automatically extract information from biomedical literature help alleviate this burden. Recently, large Language Models (LLMs), such as GPT-3 and GPT-4, have gained significant attention for their impressive performance. However, their effectiveness in BioNLP tasks and impact on method development and downstream users remain understudied. This pilot study (1) establishes the baseline performance of GPT-3 and GPT-4 at both zero-shot and one-shot settings in eight BioNLP datasets across four applications: named entity recognition, relation extraction, multi-label document classification, and semantic similarity and reasoning, (2) examines the errors produced by the LLMs and categorized the errors into three types: missingness, inconsistencies, and unwanted artificial content, and (3) provides suggestions for using LLMs in BioNLP applications. We make the datasets, baselines, and results publicly available to the community via https://github.com/qingyu-qc/gpt_bionlp_benchmark.
Authors: Shaunak Halbe, James Seale Smith, Junjiao Tian, Zsolt Kira
Abstract: In this paper, we focus on the important yet understudied problem of Continual Federated Learning (CFL), where a server communicates with a set of clients to incrementally learn new concepts over time without sharing or storing any data. The complexity of this problem is compounded by challenges from both the Continual and Federated Learning perspectives. Specifically, models trained in a CFL setup suffer from catastrophic forgetting which is exacerbated by data heterogeneity across clients. Existing attempts at this problem tend to impose large overheads on clients and communication channels or require access to stored data which renders them unsuitable for real-world use due to privacy. In this paper, we attempt to tackle forgetting and heterogeneity while minimizing overhead costs and without requiring access to any stored data. We study this problem in the context of Vision Transformers and explore parameter-efficient approaches to adapt to dynamic distributions while minimizing forgetting. We achieve this by leveraging a prompting based approach (such that only prompts and classifier heads have to be communicated) and proposing a novel and lightweight generation and distillation scheme to consolidate client models at the server. We formulate this problem for image classification and establish strong baselines for comparison, conduct experiments on CIFAR-100 as well as challenging, large-scale datasets like ImageNet-R and DomainNet. Our approach outperforms both existing methods and our own baselines by as much as 7% while significantly reducing communication and client-level computation costs. Code available at https://github.com/shaunak27/hepco-fed.
Authors: Zhuolun He, Haoyuan Wu, Xinyun Zhang, Xufeng Yao, Su Zheng, Haisheng Zheng, Bei Yu
Abstract: The integration of a complex set of Electronic Design Automation (EDA) tools to enhance interoperability is a critical concern for circuit designers. Recent advancements in large language models (LLMs) have showcased their exceptional capabilities in natural language processing and comprehension, offering a novel approach to interfacing with EDA tools. This research paper introduces ChatEDA, an autonomous agent for EDA empowered by an LLM, AutoMage, complemented by EDA tools serving as executors. ChatEDA streamlines the design flow from the Register-Transfer Level (RTL) to the Graphic Data System Version II (GDSII) by effectively managing task decomposition, script generation, and task execution. Through comprehensive experimental evaluations, ChatEDA has demonstrated its proficiency in handling diverse requirements, and our fine-tuned AutoMage model has exhibited superior performance compared to GPT-4 and other similar LLMs.
Authors: Muris Sladi\'c, Veronica Valeros, Carlos Catania, Sebastian Garcia
Abstract: Honeypots are essential tools in cybersecurity for early detection, threat intelligence gathering, and analysis of attacker's behavior. However, most of them lack the required realism to engage and fool human attackers long-term. Being easy to distinguish honeypots strongly hinders their effectiveness. This can happen because they are too deterministic, lack adaptability, or lack deepness. This work introduces shelLM, a dynamic and realistic software honeypot based on Large Language Models that generates Linux-like shell output. We designed and implemented shelLM using cloud-based LLMs. We evaluated if shelLM can generate output as expected from a real Linux shell. The evaluation was done by asking cybersecurity researchers to use the honeypot and give feedback if each answer from the honeypot was the expected one from a Linux shell. Results indicate that shelLM can create credible and dynamic answers capable of addressing the limitations of current honeypots. ShelLM reached a TNR of 0.90, convincing humans it was consistent with a real Linux shell. The source code and prompts for replicating the experiments have been publicly available.
Authors: Lei Wang, Piotr Koniusz, Tom Gedeon, Liang Zheng
Abstract: In contrastive learning, two views of an original image, generated by different augmentations, are considered a positive pair, and their similarity is required to be high. Similarly, two views of distinct images form a negative pair, with encouraged low similarity. Typically, a single similarity measure, provided by a lone projection head, evaluates positive and negative sample pairs. However, due to diverse augmentation strategies and varying intra-sample similarity, views from the same image may not always be similar. Additionally, owing to inter-sample similarity, views from different images may be more akin than those from the same image. Consequently, enforcing high similarity for positive pairs and low similarity for negative pairs may be unattainable, and in some cases, such enforcement could detrimentally impact performance. To address this challenge, we propose using multiple projection heads, each producing a distinct set of features. Our pre-training loss function emerges from a solution to the maximum likelihood estimation over head-wise posterior distributions of positive samples given observations. This loss incorporates the similarity measure over positive and negative pairs, each re-weighted by an individual adaptive temperature, regulated to prevent ill solutions. Our approach, Adaptive Multi-Head Contrastive Learning (AMCL), can be applied to and experimentally enhances several popular contrastive learning methods such as SimCLR, MoCo, and Barlow Twins. The improvement remains consistent across various backbones and linear probing epochs, and becomes more significant when employing multiple augmentation methods.
Authors: Zhuoxiao Chen, Yadan Luo, Zixin Wang, Zijian Wang, Xin Yu, Zi Huang
Abstract: LiDAR-based 3D object detection has recently seen significant advancements through active learning (AL), attaining satisfactory performance by training on a small fraction of strategically selected point clouds. However, in real-world deployments where streaming point clouds may include unknown or novel objects, the ability of current AL methods to capture such objects remains unexplored. This paper investigates a more practical and challenging research task: Open World Active Learning for 3D Object Detection (OWAL-3D), aimed at acquiring informative point clouds with new concepts. To tackle this challenge, we propose a simple yet effective strategy called Open Label Conciseness (OLC), which mines novel 3D objects with minimal annotation costs. Our empirical results show that OLC successfully adapts the 3D detection model to the open world scenario with just a single round of selection. Any generic AL policy can then be integrated with the proposed OLC to efficiently address the OWAL-3D problem. Based on this, we introduce the Open-CRB framework, which seamlessly integrates OLC with our preliminary AL method, CRB, designed specifically for 3D object detection. We develop a comprehensive codebase for easy reproducing and future research, supporting 15 baseline methods (\textit{i.e.}, active learning, out-of-distribution detection and open world detection), 2 types of modern 3D detectors (\textit{i.e.}, one-stage SECOND and two-stage PV-RCNN) and 3 benchmark 3D datasets (\textit{i.e.}, KITTI, nuScenes and Waymo). Extensive experiments evidence that the proposed Open-CRB demonstrates superiority and flexibility in recognizing both novel and known classes with very limited labeling costs, compared to state-of-the-art baselines. Source code is available at \url{https://github.com/Luoyadan/CRB-active-3Ddet/tree/Open-CRB}.
URLs: https://github.com/Luoyadan/CRB-active-3Ddet/tree/Open-CRB
Authors: Filip Ekstr\"om Kelvinius, Fredrik Lindsten
Abstract: We introduce discriminator guidance in the setting of Autoregressive Diffusion Models. The use of a discriminator to guide a diffusion process has previously been used for continuous diffusion models, and in this work we derive ways of using a discriminator together with a pretrained generative model in the discrete case. First, we show that using an optimal discriminator will correct the pretrained model and enable exact sampling from the underlying data distribution. Second, to account for the realistic scenario of using a sub-optimal discriminator, we derive a sequential Monte Carlo algorithm which iteratively takes the predictions from the discriminator into account during the generation process. We test these approaches on the task of generating molecular graphs and show how the discriminator improves the generative performance over using only the pretrained model.
Authors: Zichen Chen, Jianda Chen, Ambuj Singh, Misha Sra
Abstract: Large Language Models (LLMs) have achieved remarkable success in natural language tasks, yet understanding their reasoning processes remains a significant challenge. We address this by introducing XplainLLM, a dataset accompanying an explanation framework designed to enhance LLM transparency and reliability. Our dataset comprises 24,204 instances where each instance interprets the LLM's reasoning behavior using knowledge graphs (KGs) and graph attention networks (GAT), and includes explanations of LLMs such as the decoder-only Llama-3 and the encoder-only RoBERTa. XplainLLM also features a framework for generating grounded explanations and the debugger-scores for multidimensional quality analysis. Our explanations include why-choose and why-not-choose components, reason-elements, and debugger-scores that collectively illuminate the LLM's reasoning behavior. Our evaluations demonstrate XplainLLM's potential to reduce hallucinations and improve grounded explanation generation in LLMs. XplainLLM is a resource for researchers and practitioners to build trust and verify the reliability of LLM outputs.
Authors: Erik Derner, Kristina Batisti\v{c}, Jan Zah\'alka, Robert Babu\v{s}ka
Abstract: As large language models (LLMs) permeate more and more applications, an assessment of their associated security risks becomes increasingly necessary. The potential for exploitation by malicious actors, ranging from disinformation to data breaches and reputation damage, is substantial. This paper addresses a gap in current research by specifically focusing on security risks posed by LLMs within the prompt-based interaction scheme, which extends beyond the widely covered ethical and societal implications. Our work proposes a taxonomy of security risks along the user-model communication pipeline and categorizes the attacks by target and attack type alongside the commonly used confidentiality, integrity, and availability (CIA) triad. The taxonomy is reinforced with specific attack examples to showcase the real-world impact of these risks. Through this taxonomy, we aim to inform the development of robust and secure LLM applications, enhancing their safety and trustworthiness.
Authors: Fabian Akkerman, Peter Dieter, Martijn Mes
Abstract: Home delivery failures, traffic congestion, and relatively large handling times have a negative impact on the profitability of last-mile logistics. A potential solution is the delivery to parcel lockers or parcel shops, denoted by out-of-home (OOH) delivery. In the academic literature, models for OOH delivery were so far limited to static settings, contrasting with the sequential nature of the problem. We model the sequential decision-making problem of which OOH location to offer against what incentive for each incoming customer, taking into account future customer arrivals and choices. We propose Dynamic Selection and Pricing of OOH (DSPO), an algorithmic pipeline that uses a novel spatial-temporal state encoding as input to a convolutional neural network. We demonstrate the performance of our method by benchmarking it against two state-of-the-art approaches. Our extensive numerical study, guided by real-world data, reveals that DSPO can save 19.9%pt in costs compared to a situation without OOH locations, 7%pt compared to a static selection and pricing policy, and 3.8%pt compared to a state-of-the-art demand management benchmark. We provide comprehensive insights into the complex interplay between OOH delivery dynamics and customer behavior influenced by pricing strategies. The implications of our findings suggest practitioners to adopt dynamic selection and pricing policies.
Authors: Jiaqi Li, Yuanhao Lai, Rui Wang, Changjian Shui, Sabyasachi Sahoo, Charles X. Ling, Shichun Yang, Boyu Wang, Christian Gagn\'e, Fan Zhou
Abstract: Continual learning aims to learn a series of tasks sequentially without forgetting the knowledge acquired from the previous ones. In this work, we propose the Hessian Aware Low-Rank Perturbation algorithm for continual learning. By modeling the parameter transitions along the sequential tasks with the weight matrix transformation, we propose to apply the low-rank approximation on the task-adaptive parameters in each layer of the neural networks. Specifically, we theoretically demonstrate the quantitative relationship between the Hessian and the proposed low-rank approximation. The approximation ranks are then globally determined according to the marginal increment of the empirical loss estimated by the layer-specific gradient and low-rank approximation error. Furthermore, we control the model capacity by pruning less important parameters to diminish the parameter growth. We conduct extensive experiments on various benchmarks, including a dataset with large-scale tasks, and compare our method against some recent state-of-the-art methods to demonstrate the effectiveness and scalability of our proposed method. Empirical results show that our method performs better on different benchmarks, especially in achieving task order robustness and handling the forgetting issue. The source code is at https://github.com/lijiaqi/HALRP.
Authors: Shicheng Li, Lei Li, Shuhuai Ren, Yuanxin Liu, Yi Liu, Rundong Gao, Xu Sun, Lu Hou
Abstract: The ability to perceive how objects change over time is a crucial ingredient in human intelligence. However, current benchmarks cannot faithfully reflect the temporal understanding abilities of video-language models (VidLMs) due to the existence of static visual shortcuts. To remedy this issue, we present VITATECS, a diagnostic VIdeo-Text dAtaset for the evaluation of TEmporal Concept underStanding. Specifically, we first introduce a fine-grained taxonomy of temporal concepts in natural language in order to diagnose the capability of VidLMs to comprehend different temporal aspects. Furthermore, to disentangle the correlation between static and temporal information, we generate counterfactual video descriptions that differ from the original one only in the specified temporal aspect. We employ a semi-automatic data collection framework using large language models and human-in-the-loop annotation to obtain high-quality counterfactual descriptions efficiently. Evaluation of representative video-language understanding models confirms their deficiency in temporal understanding, revealing the need for greater emphasis on the temporal elements in video-language research.
Authors: Aditya Prakash, Ruisen Tu, Matthew Chang, Saurabh Gupta
Abstract: 3D hand pose estimation in everyday egocentric images is challenging for several reasons: poor visual signal (occlusion from the object of interaction, low resolution & motion blur), large perspective distortion (hands are close to the camera), and lack of 3D annotations outside of controlled settings. While existing methods often use hand crops as input to focus on fine-grained visual information to deal with poor visual signal, the challenges arising from perspective distortion and lack of 3D annotations in the wild have not been systematically studied. We focus on this gap and explore the impact of different practices, i.e. crops as input, incorporating camera information, auxiliary supervision, scaling up datasets. We provide several insights that are applicable to both convolutional and transformer models leading to better performance. Based on our findings, we also present WildHands, a system for 3D hand pose estimation in everyday egocentric images. Zero-shot evaluation on 4 diverse datasets (H2O, AssemblyHands, Epic-Kitchens, Ego-Exo4D) demonstrate the effectiveness of our approach across 2D and 3D metrics, where we beat past methods by 7.4% - 66%. In system level comparisons, WildHands achieves the best 3D hand pose on ARCTIC egocentric split, outperforms FrankMocap across all metrics and HaMeR on 3 out of 6 metrics while being 10x smaller and trained on 5x less data.
Authors: Aditya Prakash, Arjun Gupta, Saurabh Gupta
Abstract: Objects undergo varying amounts of perspective distortion as they move across a camera's field of view. Models for predicting 3D from a single image often work with crops around the object of interest and ignore the location of the object in the camera's field of view. We note that ignoring this location information further exaggerates the inherent ambiguity in making 3D inferences from 2D images and can prevent models from even fitting to the training data. To mitigate this ambiguity, we propose Intrinsics-Aware Positional Encoding (KPE), which incorporates information about the location of crops in the image and camera intrinsics. Experiments on three popular 3D-from-a-single-image benchmarks: depth prediction on NYU, 3D object detection on KITTI & nuScenes, and predicting 3D shapes of articulated objects on ARCTIC, show the benefits of KPE.
Authors: Jiaqi Tang, Hao Lu, Xiaogang Xu, Ruizheng Wu, Sixing Hu, Tong Zhang, Tsz Wa Cheng, Ming Ge, Ying-Cong Chen, Fugee Tsung
Abstract: Artificial Intelligence (AI)-driven defect inspection is pivotal in industrial manufacturing. Yet, many methods, tailored to specific pipelines, grapple with diverse product portfolios and evolving processes. Addressing this, we present the Incremental Unified Framework (IUF), which can reduce the feature conflict problem when continuously integrating new objects in the pipeline, making it advantageous in object-incremental learning scenarios. Employing a state-of-the-art transformer, we introduce Object-Aware Self-Attention (OASA) to delineate distinct semantic boundaries. Semantic Compression Loss (SCL) is integrated to optimize non-primary semantic space, enhancing network adaptability for novel objects. Additionally, we prioritize retaining the features of established objects during weight updates. Demonstrating prowess in both image and pixel-level defect inspection, our approach achieves state-of-the-art performance, proving indispensable for dynamic and scalable industrial inspections. Our code will be released at https://github.com/jqtangust/IUF.
Authors: Edi Sutoyo, Andrea Capiluppi
Abstract: Technical debt (TD) represents the long-term costs associated with suboptimal design or code decisions in software development, often made to meet short-term delivery goals. Self-Admitted Technical Debt (SATD) occurs when developers explicitly acknowledge these trade-offs in the codebase, typically through comments or annotations. Automated detection of SATD has become an increasingly important research area, particularly with the rise of natural language processing (NLP), machine learning (ML), and deep learning (DL) techniques that aim to streamline SATD detection. This systematic literature review provides a comprehensive analysis of SATD detection approaches published between 2014 and 2024, focusing on the evolution of techniques from NLP-based models to more advanced ML, DL, and Transformers-based models such as BERT. The review identifies key trends in SATD detection methodologies and tools, evaluates the effectiveness of different approaches using metrics like precision, recall, and F1-score, and highlights the primary challenges in this domain, including dataset heterogeneity, model generalizability, and the explainability of models. The findings suggest that while early NLP methods laid the foundation for SATD detection, more recent advancements in DL and Transformers models have significantly improved detection accuracy. However, challenges remain in scaling these models for broader industrial use. This SLR offers insights into current research gaps and provides directions for future work, aiming to improve the robustness and practicality of SATD detection tools.
Authors: Benjamin Clavi\'e
Abstract: As language-specific training data tends to be sparsely available compared to English, document retrieval in many languages has been largely relying on multilingual models. In Japanese, the best performing deep-learning based retrieval approaches rely on multilingual dense embedders, with Japanese-only models lagging far behind. However, multilingual models require considerably more compute and data to train and have higher computational and memory requirements while often missing out on culturally-relevant information. In this paper, we introduce JaColBERT, a family of multi-vector retrievers trained on two magnitudes fewer data than their multilingual counterparts while reaching competitive performance. Our strongest model largely outperform all existing monolingual Japanese retrievers on all dataset, as well as the strongest existing multilingual models on all out-of-domain tasks, highlighting the need for specialised models able to handle linguistic specificities. These results are achieved using a model with only 110 million parameters, considerably smaller than all multilingual models, and using only a limited Japanese-language. We believe our results show great promise to support Japanese retrieval-enhanced application pipelines in a wide variety of domains.
Authors: Albert Dorador
Abstract: Decades after their inception, regression forests continue to provide state-of-the-art accuracy, outperforming in this respect alternative machine learning models such as regression trees or even neural networks. However, being an ensemble method, the one aspect where regression forests tend to severely underperform regression trees is interpretability. In the present work, we revisit forest pruning, an approach that aims to have the best of both worlds: the accuracy of regression forests and the interpretability of regression trees. This pursuit, whose foundation lies at the core of random forest theory, has seen vast success in empirical studies. In this paper, we contribute theoretical results that support and qualify those empirical findings; namely, we prove the asymptotic advantage of a Lasso-pruned forest over its unpruned counterpart under extremely weak assumptions, as well as high-probability finite-sample generalization bounds for regression forests pruned according to the main methods, which we then validate by way of simulation. Then, we test the accuracy of pruned regression forests against their unpruned counterparts on 19 different datasets (16 synthetic, 3 real). We find that in the vast majority of scenarios tested, there is at least one forest-pruning method that yields equal or better accuracy than the original full forest (in expectation), while just using a small fraction of the trees. We show that, in some cases, the reduction in the size of the forest is so dramatic that the resulting sub-forest can be meaningfully merged into a single tree, obtaining a level of interpretability that is qualitatively superior to that of the original regression forest, which remains a black box.
Authors: Mengwei Xu, Wangsong Yin, Dongqi Cai, Rongjie Yi, Daliang Xu, Qipeng Wang, Bingyang Wu, Yihao Zhao, Chen Yang, Shihe Wang, Qiyang Zhang, Zhenyan Lu, Li Zhang, Shangguang Wang, Yuanchun Li, Yunxin Liu, Xin Jin, Xuanzhe Liu
Abstract: Large foundation models, including large language models (LLMs), vision transformers (ViTs), diffusion, and LLM-based multimodal models, are revolutionizing the entire machine learning lifecycle, from training to deployment. However, the substantial advancements in versatility and performance these models offer come at a significant cost in terms of hardware resources. To support the growth of these large models in a scalable and environmentally sustainable way, there has been a considerable focus on developing resource-efficient strategies. This survey delves into the critical importance of such research, examining both algorithmic and systemic aspects. It offers a comprehensive analysis and valuable insights gleaned from existing literature, encompassing a broad array of topics from cutting-edge model architectures and training/serving algorithms to practical system designs and implementations. The goal of this survey is to provide an overarching understanding of how current approaches are tackling the resource challenges posed by large foundation models and to potentially inspire future breakthroughs in this field.
Authors: Junjie Ye, Yilong Wu, Songyang Gao, Caishuang Huang, Sixian Li, Guanyu Li, Xiaoran Fan, Qi Zhang, Tao Gui, Xuanjing Huang
Abstract: Tool learning has generated widespread interest as a vital means of interaction between Large Language Models (LLMs) and the physical world. Current research predominantly emphasizes LLMs' capacity to utilize tools in well-structured environments while overlooking their stability when confronted with the inevitable noise of the real world. To bridge this gap, we introduce RoTBench, a multi-level benchmark for evaluating the robustness of LLMs in tool learning. Specifically, we establish five external environments, each featuring varying levels of noise (i.e., Clean, Slight, Medium, Heavy, and Union), providing an in-depth analysis of the model's resilience across three critical phases: tool selection, parameter identification, and content filling. Experiments involving six widely-used models underscore the urgent necessity for enhancing the robustness of LLMs in tool learning. For instance, the performance of GPT-4 even drops significantly from 80.00 to 58.10 when there is no substantial change in manual accuracy. More surprisingly, the noise correction capability inherent in the GPT family paradoxically impedes its adaptability in the face of mild noise. In light of these findings, we propose RoTTuning, a strategy that enriches the diversity of training environments to bolster the robustness of LLMs in tool learning. The code and data are available at https://github.com/Junjie-Ye/RoTBench.
Authors: Chengbo Yuan, Chuan Wen, Tong Zhang, Yang Gao
Abstract: We address the challenge of acquiring real-world manipulation skills with a scalable framework. We hold the belief that identifying an appropriate prediction target capable of leveraging large-scale datasets is crucial for achieving efficient and universal learning. Therefore, we propose to utilize 3D flow, which represents the future trajectories of 3D points on objects of interest, as an ideal prediction target. To exploit scalable data resources, we turn our attention to human videos. We develop, for the first time, a language-conditioned 3D flow prediction model directly from large-scale RGBD human video datasets. Our predicted flow offers actionable guidance, thus facilitating zero-shot skill transfer in real-world scenarios. We deploy our method with a policy based on closed-loop flow prediction. Remarkably, without any in-domain finetuning, our method achieves an impressive 81\% success rate in zero-shot human-to-robot skill transfer, covering 18 tasks in 6 scenes. Our framework features the following benefits: (1) scalability: leveraging cross-embodiment data resources; (2) wide application: multiple object categories, including rigid, articulated, and soft bodies; (3) stable skill transfer: providing actionable guidance with a small inference domain-gap. Code, data, and supplementary materials are available https://general-flow.github.io
Authors: Wesley H. Holliday, Matthew Mandelkern, Cedegao E. Zhang
Abstract: The reasoning abilities of large language models (LLMs) are the topic of a growing body of research in AI and cognitive science. In this paper, we probe the extent to which twenty-five LLMs are able to distinguish logically correct inferences from logically fallacious ones. We focus on inference patterns involving conditionals (e.g., 'If Ann has a queen, then Bob has a jack') and epistemic modals (e.g., 'Ann might have an ace', 'Bob must have a king'). These inferences have been of special interest to logicians, philosophers, and linguists, since they play a central role in the fundamental human ability to reason about distal possibilities. Assessing LLMs on these inferences is thus highly relevant to the question of how much the reasoning abilities of LLMs match those of humans. Among the LLMs we tested, all but the GPT-4 model family often make basic mistakes with conditionals, though zero-shot chain-of-thought prompting helps them make fewer mistakes. Moreover, even the GPT-4 family displays logically inconsistent judgments across inference patterns involving epistemic modals, and almost all models give answers to certain complex conditional inferences widely discussed in the literature that do not match human judgments. These results highlight gaps in basic logical reasoning in today's LLMs.
Authors: V. K. Cody Bumgardner, Mitchell A. Klusty, W. Vaiden Logan, Samuel E. Armstrong, Caylin Hickey, Jeff Talbert
Abstract: This paper introduces a user-friendly platform developed by the University of Kentucky Center for Applied AI, designed to make large, customized language models (LLMs) more accessible. By capitalizing on recent advancements in multi-LoRA inference, the system efficiently accommodates custom adapters for a diverse range of users and projects. The paper outlines the system's architecture and key features, encompassing dataset curation, model training, secure inference, and text-based feature extraction. We illustrate the establishment of a tenant-aware computational network using agent-based methods, securely utilizing islands of isolated resources as a unified system. The platform strives to deliver secure LLM services, emphasizing process and data isolation, end-to-end encryption, and role-based resource authentication. This contribution aligns with the overarching goal of enabling simplified access to cutting-edge AI models and technology in support of scientific discovery.
Authors: Wolf Rieder, Philip Raschke, Thomas Cory
Abstract: The World Wide Web's connectivity is greatly attributed to the HTTP protocol, with HTTP messages offering informative header fields that appeal to disciplines like web security and privacy, especially concerning web tracking. Despite existing research employing HTTP request messages to identify web trackers, HTTP response headers are often overlooked. This study endeavors to design effective machine learning classifiers for web tracker detection using binarized HTTP response headers. Data from the Chrome, Firefox, and Brave browsers, obtained through the traffic monitoring browser extension T.EX, serves as our dataset. Ten supervised models were trained on Chrome data and tested across all browsers, including a Chrome dataset from a year later. The results demonstrated high accuracy, F1-score, precision, recall, and minimal log-loss error for Chrome and Firefox, but subpar performance on Brave, potentially due to its distinct data distribution and feature set. The research suggests that these classifiers are viable for web tracker detection. However, real-world application testing remains pending, and the distinction between tracker types and broader label sources could be explored in future studies.
Authors: Yunqing Bao, Bin Hu
Abstract: This paper presents a novel algorithm for non-destructive damage detection for steel ropes in high-altitude environments (aerial ropeway). The algorithm comprises two key components: First, a segmentation model named RGBD-UNet is designed to accurately extract steel ropes from complex backgrounds. This model is equipped with the capability to process and combine color and depth information through the proposed CMA module. Second, a detection model named VovNetV3.5 is developed to differentiate between normal and abnormal steel ropes. It integrates the VovNet architecture with a DBB module to enhance performance. Besides, a novel background augmentation method is proposed to enhance the generalization ability of the segmentation model. Datasets containing images of steel ropes in different scenarios are created for the training and testing of both the segmentation and detection models. Experiments demonstrate a significant improvement over baseline models. On the proposed dataset, the highest accuracy achieved by the detection model reached 0.975, and the maximum F-measure achieved by the segmentation model reached 0.948.
Authors: Ziqiao Shang, Bin Liu, Fengmao Lv, Fei Teng, Tianrui Li
Abstract: Facial action unit (AU) detection has long encountered the challenge of detecting subtle feature differences when AUs activate. Existing methods often rely on encoding pixel-level information of AUs, which not only encodes additional redundant information but also leads to increased model complexity and limited generalizability. Additionally, the accuracy of AU detection is negatively impacted by the class imbalance issue of each AU type, and the presence of noisy and false AU labels. In this paper, we introduce a novel contrastive learning framework aimed for AU detection that incorporates both self-supervised and supervised signals, thereby enhancing the learning of discriminative features for accurate AU detection. To tackle the class imbalance issue, we employ a negative sample re-weighting strategy that adjusts the step size of updating parameters for minority and majority class samples. Moreover, to address the challenges posed by noisy and false AU labels, we employ a sampling technique that encompasses three distinct types of positive sample pairs. This enables us to inject self-supervised signals into the supervised signal, effectively mitigating the adverse effects of noisy labels. Our experimental assessments, conducted on four widely-utilized benchmark datasets (BP4D, DISFA, GFT and Aff-Wild2), underscore the superior performance of our approach compared to state-of-the-art methods of AU detection. Our code is available at \url{https://github.com/Ziqiao-Shang/AUNCE}.
Authors: Khiem Phi, Noushin Salek Faramarzi, Chenlu Wang, Ritwik Banerjee
Abstract: Whataboutism, a potent tool for disrupting narratives and sowing distrust, remains under-explored in quantitative NLP research. Moreover, past work has not distinguished its use as a strategy for misinformation and propaganda from its use as a tool for pragmatic and semantic framing. We introduce new datasets from Twitter and YouTube, revealing overlaps as well as distinctions between whataboutism, propaganda, and the tu quoque fallacy. Furthermore, drawing on recent work in linguistic semantics, we differentiate the `what about' lexical construct from whataboutism. Our experiments bring to light unique challenges in its accurate detection, prompting the introduction of a novel method using attention weights for negative sample mining. We report significant improvements of 4% and 10% over previous state-of-the-art methods in our Twitter and YouTube collections, respectively.
Authors: Kim Hammar, Tao Li, Rolf Stadler, Quanyan Zhu
Abstract: We study automated security response for an IT infrastructure and formulate the interaction between an attacker and a defender as a partially observed, non-stationary game. We relax the standard assumption that the game model is correctly specified and consider that each player has a probabilistic conjecture about the model, which may be misspecified in the sense that the true model has probability 0. This formulation allows us to capture uncertainty and misconception about the infrastructure and the intents of the players. To learn effective game strategies online, we design Conjectural Online Learning (COL), a novel method where a player iteratively adapts its conjecture using Bayesian learning and updates its strategy through rollout. We prove that the conjectures converge to best fits, and we provide a bound on the performance improvement that rollout enables with a conjectured model. To characterize the steady state of the game, we propose a variant of the Berk-Nash equilibrium. We present COL through an advanced persistent threat use case. Testbed evaluations show that COL produces effective security strategies that adapt to a changing environment. We also find that COL enables faster convergence than current reinforcement learning techniques.
Authors: Zefeng Wang, Zhen Han, Shuo Chen, Fan Xue, Zifeng Ding, Xun Xiao, Volker Tresp, Philip Torr, Jindong Gu
Abstract: Multimodal LLMs (MLLMs) with a great ability of text and image understanding have received great attention. To achieve better reasoning with MLLMs, Chain-of-Thought (CoT) reasoning has been widely explored, which further promotes MLLMs' explainability by giving intermediate reasoning steps. Despite the strong power demonstrated by MLLMs in multimodal reasoning, recent studies show that MLLMs still suffer from adversarial images. This raises the following open questions: Does CoT also enhance the adversarial robustness of MLLMs? What do the intermediate reasoning steps of CoT entail under adversarial attacks? To answer these questions, we first generalize existing attacks to CoT-based inferences by attacking the two main components, i.e., rationale and answer. We find that CoT indeed improves MLLMs' adversarial robustness against the existing attack methods by leveraging the multi-step reasoning process, but not substantially. Based on our findings, we further propose a novel attack method, termed as stop-reasoning attack, that attacks the model while bypassing the CoT reasoning process. Experiments on three MLLMs and two visual reasoning datasets verify the effectiveness of our proposed method. We show that stop-reasoning attack can result in misled predictions and outperform baseline attacks by a significant margin.
Authors: Derong Xu, Ziheng Zhang, Zhihong Zhu, Zhenxi Lin, Qidong Liu, Xian Wu, Tong Xu, Wanyu Wang, Yuyang Ye, Xiangyu Zhao, Enhong Chen, Yefeng Zheng
Abstract: Model editing aims to precisely alter the behaviors of large language models (LLMs) in relation to specific knowledge, while leaving unrelated knowledge intact. This approach has proven effective in addressing issues of hallucination and outdated information in LLMs. However, the potential of using model editing to modify knowledge in the medical field remains largely unexplored, even though resolving hallucination is a pressing need in this area. Our observations indicate that current methods face significant challenges in dealing with specialized and complex knowledge in medical domain. Therefore, we propose MedLaSA, a novel Layer-wise Scalable Adapter strategy for medical model editing. MedLaSA harnesses the strengths of both adding extra parameters and locate-then-edit methods for medical model editing. We utilize causal tracing to identify the association of knowledge in neurons across different layers, and generate a corresponding scale set from the association value for each piece of knowledge. Subsequently, we incorporate scalable adapters into the dense layers of LLMs. These adapters are assigned scaling values based on the corresponding specific knowledge, which allows for the adjustment of the adapter's weight and rank. The more similar the content, the more consistent the scale between them. This ensures precise editing of semantically identical knowledge while avoiding impact on unrelated knowledge. To evaluate the editing impact on the behaviours of LLMs, we propose two model editing studies for medical domain: (1) editing factual knowledge for medical specialization and (2) editing the explanatory ability for complex knowledge. We build two novel medical benchmarking datasets and introduce a series of challenging and comprehensive metrics. Extensive experiments on medical LLMs demonstrate the editing efficiency of MedLaSA, without affecting unrelated knowledge.
Authors: Zhitao He, Pengfei Cao, Chenhao Wang, Zhuoran Jin, Yubo Chen, Jiexin Xu, Huaijun Li, Xiaojian Jiang, Kang Liu, Jun Zhao
Abstract: With the development of deep learning, natural language processing technology has effectively improved the efficiency of various aspects of the traditional judicial industry. However, most current efforts focus on tasks within individual judicial stages, making it difficult to handle complex tasks that span multiple stages. As the autonomous agents powered by large language models are becoming increasingly smart and able to make complex decisions in real-world settings, offering new insights for judicial intelligence. In this paper, (1) we propose a novel multi-agent framework, AgentsCourt, for judicial decision-making. Our framework follows the classic court trial process, consisting of court debate simulation, legal resources retrieval and decision-making refinement to simulate the decision-making of judge. (2) we introduce SimuCourt, a judicial benchmark that encompasses 420 Chinese judgment documents, spanning the three most common types of judicial cases. Furthermore, to support this task, we construct a large-scale legal knowledge base, Legal-KB, with multi-resource legal knowledge. (3) Extensive experiments show that our framework outperforms the existing advanced methods in various aspects, especially in generating legal articles, where our model achieves significant improvements of 8.6% and 9.1% F1 score in the first and second instance settings, respectively.
Authors: Zeyu Zhang, Khandaker Asif Ahmed, Md Rakibul Hasan, Tom Gedeon, Md Zakir Hossain
Abstract: Diabetes, resulting from inadequate insulin production or utilization, causes extensive harm to the body. Existing diagnostic methods are often invasive and come with drawbacks, such as cost constraints. Although there are machine learning models like Classwise k Nearest Neighbor (CkNN) and General Regression Neural Network (GRNN), they struggle with imbalanced data and result in under-performance. Leveraging advancements in sensor technology and machine learning, we propose a non-invasive diabetes diagnosis using a Back Propagation Neural Network (BPNN) with batch normalization, incorporating data re-sampling and normalization for class balancing. Our method addresses existing challenges such as limited performance associated with traditional machine learning. Experimental results on three datasets show significant improvements in overall accuracy, sensitivity, and specificity compared to traditional methods. Notably, we achieve accuracies of 89.81% in Pima diabetes dataset, 75.49% in CDC BRFSS2015 dataset, and 95.28% in Mesra Diabetes dataset. This underscores the potential of deep learning models for robust diabetes diagnosis. See project website https://steve-zeyu-zhang.github.io/DiabetesDiagnosis/
Authors: Tianshuo Peng, Zuchao Li, Lefei Zhang, Hai Zhao, Ping Wang, Bo Du
Abstract: Large Language Models (LLMs), benefiting from the auto-regressive modelling approach performed on massive unannotated texts corpora, demonstrates powerful perceptual and reasoning capabilities. However, as for extending auto-regressive modelling to multi-modal scenarios to build Large Multi-modal Models (LMMs), there lies a great difficulty that the image information is processed in the LMM as continuous visual embeddings, which cannot obtain discrete supervised labels for classification.In this paper, we successfully perform multi-modal auto-regressive modeling with a unified objective for the first time.Specifically, we propose the concept of visual tokens, which maps the visual features to probability distributions over LLM's vocabulary, providing supervision information for visual modelling.We further explore the distribution of visual features in the semantic space within LMM and the possibility of using text embeddings to represent visual information.Experimental results and ablation studies on 5 VQA tasks and 4 benchmark toolkits validate the powerful performance of our proposed approach.
Authors: Shuangjian Li, Tao Zhu, Mingxing Nie, Huansheng Ning, Zhenyu Liu, Liming Chen
Abstract: Traditional deep learning methods struggle to simultaneously segment, recognize, and forecast human activities from sensor data. This limits their usefulness in many fields such as healthcare and assisted living, where real-time understanding of ongoing and upcoming activities is crucial. This paper introduces P2LHAP, a novel Patch-to-Label Seq2Seq framework that tackles all three tasks in a efficient single-task model. P2LHAP divides sensor data streams into a sequence of "patches", served as input tokens, and outputs a sequence of patch-level activity labels including the predicted future activities. A unique smoothing technique based on surrounding patch labels, is proposed to identify activity boundaries accurately. Additionally, P2LHAP learns patch-level representation by sensor signal channel-independent Transformer encoders and decoders. All channels share embedding and Transformer weights across all sequences. Evaluated on three public datasets, P2LHAP significantly outperforms the state-of-the-art in all three tasks, demonstrating its effectiveness and potential for real-world applications.
Authors: Jeongsol Kim, Geon Yeong Park, Jong Chul Ye
Abstract: Reverse sampling and score-distillation have emerged as main workhorses in recent years for image manipulation using latent diffusion models (LDMs). While reverse diffusion sampling often requires adjustments of LDM architecture or feature engineering, score distillation offers a simple yet powerful model-agnostic approach, but it is often prone to mode-collapsing. To address these limitations and leverage the strengths of both approaches, here we introduce a novel framework called {\em DreamSampler}, which seamlessly integrates these two distinct approaches through the lens of regularized latent optimization. Similar to score-distillation, DreamSampler is a model-agnostic approach applicable to any LDM architecture, but it allows both distillation and reverse sampling with additional guidance for image editing and reconstruction. Through experiments involving image editing, SVG reconstruction and etc, we demonstrate the competitive performance of DreamSampler compared to existing approaches, while providing new applications. Code: https://github.com/DreamSampler/dream-sampler
Authors: Chunhang Zheng, Kechao Cai
Abstract: Traditional approaches to semantic communication tasks rely on the knowledge of the signal-to-noise ratio (SNR) to mitigate channel noise. Moreover, these methods necessitate training under specific SNR conditions, entailing considerable time and computational resources. In this paper, we propose GeNet, a Graph Neural Network (GNN)-based paradigm for semantic communication aimed at combating noise, thereby facilitating Task-Oriented Communication (TOC). We propose a novel approach where we first transform the input data image into graph structures. Then we leverage a GNN-based encoder to extract semantic information from the source data. This extracted semantic information is then transmitted through the channel. At the receiver's end, a GNN-based decoder is utilized to reconstruct the relevant semantic information from the source data for TOC. Through experimental evaluation, we show GeNet's effectiveness in anti-noise TOC while decoupling the SNR dependency. We further evaluate GeNet's performance by varying the number of nodes, revealing its versatility as a new paradigm for semantic communication. Additionally, we show GeNet's robustness to geometric transformations by testing it with different rotation angles, without resorting to data augmentation.
Authors: Md. Simul Hasan Talukder, Sharmin Akter, Abdullah Hafez Nur, Mohammad Aljaidi, Rejwan Bin Sulaiman
Abstract: Sugarcane, a key crop for the world's sugar industry, is prone to several diseases that have a substantial negative influence on both its yield and quality. To effectively manage and implement preventative initiatives, diseases must be detected promptly and accurately. In this study, we present a unique model called sugarcaneNet2024 that outperforms previous methods for automatically and quickly detecting sugarcane disease through leaf image processing. Our proposed model consolidates an optimized weighted average ensemble of seven customized and LASSO-regularized pre-trained models, particularly InceptionV3, InceptionResNetV2, DenseNet201, DenseNet169, Xception, and ResNet152V2. Initially, we added three more dense layers with 0.0001 LASSO regularization, three 30% dropout layers, and three batch normalizations with renorm enabled at the bottom of these pre-trained models to improve the performance. The accuracy of sugarcane leaf disease classification was greatly increased by this addition. Following this, several comparative studies between the average ensemble and individual models were carried out, indicating that the ensemble technique performed better. The average ensemble of all modified pre-trained models produced outstanding outcomes: 100%, 99%, 99%, and 99.45% for f1 score, precision, recall, and accuracy, respectively. Performance was further enhanced by the implementation of an optimized weighted average ensemble technique incorporated with grid search. This optimized sugarcaneNet2024 model performed the best for detecting sugarcane diseases, having achieved accuracy, precision, recall, and F1 score of 99.67%, 100%, 100%, and 100% , respectively.
Authors: Yuji Cao, Huan Zhao, Yuheng Cheng, Ting Shu, Yue Chen, Guolong Liu, Gaoqi Liang, Junhua Zhao, Jinyue Yan, Yun Li
Abstract: With extensive pre-trained knowledge and high-level general capabilities, large language models (LLMs) emerge as a promising avenue to augment reinforcement learning (RL) in aspects such as multi-task learning, sample efficiency, and high-level task planning. In this survey, we provide a comprehensive review of the existing literature in LLM-enhanced RL and summarize its characteristics compared to conventional RL methods, aiming to clarify the research scope and directions for future studies. Utilizing the classical agent-environment interaction paradigm, we propose a structured taxonomy to systematically categorize LLMs' functionalities in RL, including four roles: information processor, reward designer, decision-maker, and generator. For each role, we summarize the methodologies, analyze the specific RL challenges that are mitigated, and provide insights into future directions. Lastly, a comparative analysis of each role, potential applications, prospective opportunities, and challenges of the LLM-enhanced RL are discussed. By proposing this taxonomy, we aim to provide a framework for researchers to effectively leverage LLMs in the RL field, potentially accelerating RL applications in complex applications such as robotics, autonomous driving, and energy systems.
Authors: Shujie Hu, Long Zhou, Shujie Liu, Sanyuan Chen, Lingwei Meng, Hongkun Hao, Jing Pan, Xunying Liu, Jinyu Li, Sunit Sivasankaran, Linquan Liu, Furu Wei
Abstract: The recent advancements in large language models (LLMs) have revolutionized the field of natural language processing, progressively broadening their scope to multimodal perception and generation. However, effectively integrating listening capabilities into LLMs poses significant challenges, particularly with respect to generalizing across varied contexts and executing complex auditory tasks. In this work, we introduce WavLLM, a robust and adaptive speech large language model with dual encoders, and a prompt-aware LoRA weight adapter, optimized by a two-stage curriculum learning approach. Leveraging dual encoders, we decouple different types of speech information, utilizing a Whisper encoder to process the semantic content of speech, and a WavLM encoder to capture the unique characteristics of the speaker's identity. Within the curriculum learning framework, WavLLM first builds its foundational capabilities by optimizing on mixed elementary single tasks, followed by advanced multi-task training on more complex tasks such as combinations of the elementary tasks. To enhance the flexibility and adherence to different tasks and instructions, a prompt-aware LoRA weight adapter is introduced in the second advanced multi-task training stage. We validate the proposed model on universal speech benchmarks including tasks such as ASR, ST, SV, ER, and also apply it to specialized datasets like Gaokao English listening comprehension set for SQA, and speech Chain-of-Thought (CoT) evaluation set. Experiments demonstrate that the proposed model achieves state-of-the-art performance across a range of speech tasks on the same model size, exhibiting robust generalization capabilities in executing complex tasks using CoT approach. Furthermore, our model successfully completes Gaokao tasks without specialized training. The codes, models, audio, and Gaokao evaluation set can be accessed at \url{aka.ms/wavllm}.
Authors: Sandeep Nagar, Ehsan Farahbakhsh, Joseph Awange, Rohitash Chandra
Abstract: Supervised machine learning methods for geological mapping via remote sensing face limitations due to the scarcity of accurately labelled training data that can be addressed by unsupervised learning, such as dimensionality reduction and clustering. Dimensionality reduction methods have the potential to play a crucial role in improving the accuracy of geological maps. Although conventional dimensionality reduction methods may struggle with nonlinear data, unsupervised deep learning models such as autoencoders can model non-linear relationships. Stacked autoencoders feature multiple interconnected layers to capture hierarchical data representations useful for remote sensing data. We present an unsupervised machine learning-based framework for processing remote sensing data using stacked autoencoders for dimensionality reduction and k-means clustering for mapping geological units. We use Landsat 8, ASTER, and Sentinel-2 datasets to evaluate the framework for geological mapping of the Mutawintji region in Western New South Wales, Australia. We also compare stacked autoencoders with principal component analysis (PCA) and canonical autoencoders. Our results reveal that the framework produces accurate and interpretable geological maps, efficiently discriminating rock units. The results reveal that the combination of stacked autoencoders with Sentinel-2 data yields the best performance accuracy when compared to other combinations. We find that stacked autoencoders enable better extraction of complex and hierarchical representations of the input data when compared to canonical autoencoders and PCA. We also find that the generated maps align with prior geological knowledge of the study area while providing novel insights into geological structures.
Authors: Lei Zhang, Kaixin Bai, Guowen Huang, Zhenshan Bing, Zhaopeng Chen, Alois Knoll, Jianwei Zhang
Abstract: The deep learning models has significantly advanced dexterous manipulation techniques for multi-fingered hand grasping. However, the contact information-guided grasping in cluttered environments remains largely underexplored. To address this gap, we have developed a method for generating multi-fingered hand grasp samples in cluttered settings through contact semantic map. We introduce a contact semantic conditional variational autoencoder network (CoSe-CVAE) for creating comprehensive contact semantic map from object point cloud. We utilize grasp detection method to estimate hand grasp poses from the contact semantic map. Finally, an unified grasp evaluation model is designed to assess grasp quality and collision probability, substantially improving the reliability of identifying optimal grasps in cluttered scenarios. Our grasp generation method has demonstrated remarkable success, outperforming state-of-the-art methods by at least 4.65% with 81.0% average grasping success rate in real-world single-object environment and 75.3% grasping success rate in cluttered scenes. We also proposed the multi-modal multi-fingered grasping dataset generation method. Our multi-fingered hand grasping dataset outperforms previous datasets in scene diversity, modality diversity. The dataset, code and supplementary materials can be found at https://sites.google.com/view/ffh-cluttered-grasping.
Authors: Juhwan Choi, Jungmin Yun, Kyohoon Jin, YoungBin Kim
Abstract: The quality of the dataset is crucial for ensuring optimal performance and reliability of downstream task models. However, datasets often contain noisy data inadvertently included during the construction process. Numerous attempts have been made to correct this issue through human annotators. However, hiring and managing human annotators is expensive and time-consuming. As an alternative, recent studies are exploring the use of large language models (LLMs) for data annotation. In this study, we present a case study that extends the application of LLM-based data annotation to enhance the quality of existing datasets through a cleansing strategy. Specifically, we leverage approaches such as chain-of-thought and majority voting to imitate human annotation and classify unrelated documents from the Multi-News dataset, which is widely used for the multi-document summarization task. Through our proposed cleansing method, we introduce an enhanced Multi-News+. By employing LLMs for data cleansing, we demonstrate an efficient and effective approach to improving dataset quality without relying on expensive human annotation efforts.
Authors: Jo\~ao Gabriel Vinholi, Marco Chini, Anis Amziane, Renato Machado, Danilo Silva, Patrick Matgen
Abstract: Comparing images captured by disparate sensors is a common challenge in remote sensing. This requires image translation -- converting imagery from one sensor domain to another while preserving the original content. Denoising Diffusion Implicit Models (DDIM) are potential state-of-the-art solutions for such domain translation due to their proven superiority in multiple image-to-image translation tasks in classic computer vision. However, these models struggle with large-scale multi-patch imagery, often focusing solely on small patches and resulting in inconsistencies across the full image. To overcome these limitations, we propose a novel method that leverages DDIM for effective optical image translation over large areas. Our approach super-resolves large-scale low spatial resolution images into high-resolution equivalents from disparate optical sensors, ensuring uniformity across hundreds of patches. Our contributions lie in new forward and reverse diffusion processes, which are crucial for addressing the challenges of large-scale image translation. Extensive experiments with a dataset of paired Sentinel-II and Planet Dove images demonstrate that our approach provides precise domain adaptation and artifact reduction. The proposed method preserves image content while improving radiometric (color) accuracy and feature representation. The outcome is a high-resolution large-scale image with consistent patches, vital for applications such as heterogeneous change detection (HCD). We present a unique training and testing algorithm rooted in DDIMs, a thorough image quality assessment, and a comparative study against the standard classifier-free guided DDIM framework and five other leading methods. The usefulness of our approach is further demonstrated in HCD tasks performed in the urban settings of Beirut, Lebanon, and Austin, USA.
Authors: Peihan Tu, Li-Yi Wei, Matthias Zwicker
Abstract: Texture plays a vital role in enhancing visual richness in both real photographs and computer-generated imagery. However, the process of editing textures often involves laborious and repetitive manual adjustments of textons, which are the recurring local patterns that characterize textures. This work introduces a fully unsupervised approach for representing textures using a compositional neural model that captures individual textons. We represent each texton as a 2D Gaussian function whose spatial support approximates its shape, and an associated feature that encodes its detailed appearance. By modeling a texture as a discrete composition of Gaussian textons, the representation offers both expressiveness and ease of editing. Textures can be edited by modifying the compositional Gaussians within the latent space, and new textures can be efficiently synthesized by feeding the modified Gaussians through a generator network in a feed-forward manner. This approach enables a wide range of applications, including transferring appearance from an image texture to another image, diversifying textures,texture interpolation, revealing/modifying texture variations, edit propagation, texture animation, and direct texton manipulation. The proposed approach contributes to advancing texture analysis, modeling, and editing techniques, and opens up new possibilities for creating visually appealing images with controllable textures.
Authors: Hai Siong Tan, Kuancheng Wang, Rafe Mcbeth
Abstract: In this work, we present a novel application of an uncertainty-quantification framework called Deep Evidential Learning in the domain of radiotherapy dose prediction. Using medical images of the Open Knowledge-Based Planning Challenge dataset, we found that this model can be effectively harnessed to yield uncertainty estimates that inherited correlations with prediction errors upon completion of network training. This was achieved only after reformulating the original loss function for a stable implementation. We found that (i)epistemic uncertainty was highly correlated with prediction errors, with various association indices comparable or stronger than those for Monte-Carlo Dropout and Deep Ensemble methods, (ii)the median error varied with uncertainty threshold much more linearly for epistemic uncertainty in Deep Evidential Learning relative to these other two conventional frameworks, indicative of a more uniformly calibrated sensitivity to model errors, (iii)relative to epistemic uncertainty, aleatoric uncertainty demonstrated a more significant shift in its distribution in response to Gaussian noise added to CT intensity, compatible with its interpretation as reflecting data noise. Collectively, our results suggest that Deep Evidential Learning is a promising approach that can endow deep-learning models in radiotherapy dose prediction with statistical robustness. Towards enhancing its clinical relevance, we demonstrate how we can use such a model to construct the predicted Dose-Volume-Histograms' confidence intervals.
Authors: Justin Engelmann, Miguel O. Bernabeu
Abstract: Artificial Intelligence in medicine is traditionally limited by the lack of massive training datasets. Foundation models, pre-trained models that can be adapted to downstream tasks with small datasets, could alleviate this problem. Researchers at Moorfields Eye Hospital (MEH) proposed RETFound-MEH, a retinal foundation model trained on 900,000 images, including private hospital data. Recently, data-efficient DERETFound was proposed providing comparable performance while being trained on only 150,000 publicly available images. However, both these models required very substantial resources to train initially and are resource-intensive in downstream use. We propose a novel Token Reconstruction objective that we use to train RETFound-Green, a retinal foundation model trained using only 75,000 publicly available images and 400 times less compute. We estimate the cost of training RETFound-MEH and DERETFound at \$10,000 and \$14,000, respectively. RETFound-Green could be trained for less than \$100, with equally reduced environmental impact. RETFound-Green is also far more efficient in downstream use: it can be downloaded 14 times faster, computes vector embeddings 2.7 times faster which then require 2.6 times less storage space. Despite this, RETFound-Green does not perform systematically worse. In fact, on various task on three downstream datasets from Brazil, India and China, it performs best on 68 tasks out of 119 comparisons, versus 21 for DERETFound and 13 for RETFound-MEH. Our results suggest that RETFound-Green is a very efficient, high-performance retinal foundation model. We anticipate that our Token Reconstruction objective could be scaled up for even higher performance and be applied to other domains beyond retinal imaging.
Authors: Juhwan Choi, Yeonghwa Kim, Seunguk Yu, JungMin Yun, YoungBin Kim
Abstract: Although pre-trained language models have exhibited great flexibility and versatility with prompt-based few-shot learning, they suffer from the extensive parameter size and limited applicability for inference. Recent studies have suggested that PLMs be used as dataset generators and a tiny task-specific model be trained to achieve efficient inference. However, their applicability to various domains is limited because they tend to generate domain-specific datasets. In this work, we propose a novel approach to universal domain generalization that generates a dataset regardless of the target domain. This allows for generalization of the tiny task model to any domain that shares the label space, thus enhancing the real-world applicability of the dataset generation paradigm. Our experiments indicate that the proposed method accomplishes generalizability across various domains while using a parameter set that is orders of magnitude smaller than PLMs.
Authors: Yu Pan, Yuguang Yang, Heng Lu, Lei Ma, Jianjun Zhao
Abstract: The continuous evolution of pre-trained speech models has greatly advanced Speech Emotion Recognition (SER). However, current research typically relies on utterance-level emotion labels, inadequately capturing the complexity of emotions within a single utterance. In this paper, we introduce GMP-TL, a novel SER framework that employs gender-augmented multi-scale pseudo-label (GMP) based transfer learning to mitigate this gap. Specifically, GMP-TL initially uses the pre-trained HuBERT, implementing multi-task learning and multi-scale k-means clustering to acquire frame-level GMPs. Subsequently, to fully leverage frame-level GMPs and utterance-level emotion labels, a two-stage model fine-tuning approach is presented to further optimize GMP-TL. Experiments on IEMOCAP show that our GMP-TL attains a WAR of 80.0% and an UAR of 82.0%, achieving superior performance compared to state-of-the-art unimodal SER methods while also yielding comparable results to multimodal SER approaches.
Authors: Pranab Sahoo, Prabhash Meharia, Akash Ghosh, Sriparna Saha, Vinija Jain, Aman Chadha
Abstract: The rapid advancement of foundation models (FMs) across language, image, audio, and video domains has shown remarkable capabilities in diverse tasks. However, the proliferation of FMs brings forth a critical challenge: the potential to generate hallucinated outputs, particularly in high-stakes applications. The tendency of foundation models to produce hallucinated content arguably represents the biggest hindrance to their widespread adoption in real-world scenarios, especially in domains where reliability and accuracy are paramount. This survey paper presents a comprehensive overview of recent developments that aim to identify and mitigate the problem of hallucination in FMs, spanning text, image, video, and audio modalities. By synthesizing recent advancements in detecting and mitigating hallucination across various modalities, the paper aims to provide valuable insights for researchers, developers, and practitioners. Essentially, it establishes a clear framework encompassing definition, taxonomy, and detection strategies for addressing hallucination in multimodal foundation models, laying the foundation for future research in this pivotal area.
Authors: Dan Biderman, Jacob Portes, Jose Javier Gonzalez Ortiz, Mansheej Paul, Philip Greengard, Connor Jennings, Daniel King, Sam Havens, Vitaliy Chiley, Jonathan Frankle, Cody Blakeney, John P. Cunningham
Abstract: Low-Rank Adaptation (LoRA) is a widely-used parameter-efficient finetuning method for large language models. LoRA saves memory by training only low rank perturbations to selected weight matrices. In this work, we compare the performance of LoRA and full finetuning on two target domains, programming and mathematics. We consider both the instruction finetuning (approximately 100K prompt-response pairs) and continued pretraining (20B unstructured tokens) data regimes. Our results show that, in the standard low-rank settings, LoRA substantially underperforms full finetuning. Nevertheless, LoRA better maintains the base model's performance on tasks outside the target domain. We show that LoRA mitigates forgetting more than common regularization techniques such as weight decay and dropout; it also helps maintain more diverse generations. Finally, we show that full finetuning learns perturbations with a rank that is 10-100X greater than typical LoRA configurations, possibly explaining some of the reported gaps. We conclude by proposing best practices for finetuning with LoRA.
Authors: Virgile Rennard, Guokan Shang, Michalis Vazirgiannis, Julie Hunter
Abstract: We introduce an extractive summarization system for meetings that leverages discourse structure to better identify salient information from complex multi-party discussions. Using discourse graphs to represent semantic relations between the contents of utterances in a meeting, we train a GNN-based node classification model to select the most important utterances, which are then combined to create an extractive summary. Experimental results on AMI and ICSI demonstrate that our approach surpasses existing text-based and graph-based extractive summarization systems, as measured by both classification and summarization metrics. Additionally, we conduct ablation studies on discourse structure and relation type to provide insights for future NLP applications leveraging discourse analysis theory.
Authors: Daniel Chin, Yuxuan Wang, Gus Xia
Abstract: Large Language Model (LLM) -in-the-loop applications have been shown to effectively interpret the human user's commands, make plans, and operate external tools/systems accordingly. Still, the operation scope of the LLM agent is limited to passively following the user, requiring the user to frame his/her needs with regard to the underlying tools/systems. We note that the potential of an LLM-Agent User Interface (LAUI) is much greater. A user mostly ignorant to the underlying tools/systems should be able to work with a LAUI to discover an emergent workflow. Contrary to the conventional way of designing an explorable GUI to teach the user a predefined set of ways to use the system, in the ideal LAUI, the LLM agent is initialized to be proficient with the system, proactively studies the user and his/her needs, and proposes new interaction schemes to the user. To illustrate LAUI, we present Flute X GPT, a concrete example using an LLM agent, a prompt manager, and a flute-tutoring multi-modal software-hardware system to facilitate the complex, real-time user experience of learning to play the flute.
Authors: Ruolin Wang, Yuejiao Xu, Jianmin Ji
Abstract: Formal representations of traffic scenarios can be used to generate test cases for the safety verification of autonomous driving. However, most existing methods are limited to highway or highly simplified intersection scenarios due to the intricacy and diversity of traffic scenarios. In response, we propose Traffic Scenario Logic (TSL), which is a spatial-temporal logic designed for modeling and reasoning of urban pedestrian-free traffic scenarios. TSL provides a formal representation of the urban road network that can be derived from OpenDRIVE, i.e., the de facto industry standard of high-definition maps for autonomous driving, enabling the representation of a broad range of traffic scenarios without discretization approximations. We implemented the reasoning of TSL using Telingo, i.e., a solver for temporal programs based on the Answer Set Programming, and tested it on different urban road layouts. Demonstrations show the effectiveness of TSL in test scenario generation and its potential value in areas like decision-making and control verification of autonomous driving. The code for TSL reasoning is opened.
Authors: Aditya Gunturu, Shivesh Jadon, Nandi Zhang, Morteza Faraji, Jarin Thundathil, Tafreed Ahmad, Wesley Willett, Ryo Suzuki
Abstract: Large Language Models (LLMs) are gaining popularity as tools for reading and summarization aids. However, little is known about their potential benefits when integrated with mixed reality (MR) interfaces to support everyday reading assistants. We developed RealitySummary, an MR reading assistant that seamlessly integrates LLMs with always-on camera access, OCR-based text extraction, and augmented spatial and visual responses in MR interfaces. Developed iteratively, RealitySummary evolved across three versions, each shaped by user feedback and reflective analysis: 1) a preliminary user study to understand user perceptions (N=12), 2) an in-the-wild deployment to explore real-world usage (N=11), and 3) a diary study to capture insights from real-world work contexts (N=5). Our findings highlight the unique advantages of combining AI and MR, including an always-on implicit assistant, minimal context switching, and spatial affordances, demonstrating significant potential for future LLM-MR interfaces beyond traditional screen-based interactions.
Authors: Thong Thanh Nguyen, Zhiyuan Hu, Xiaobao Wu, Cong-Duy T Nguyen, See-Kiong Ng, Anh Tuan Luu
Abstract: Seeking answers effectively for long videos is essential to build video question answering (videoQA) systems. Previous methods adaptively select frames and regions from long videos to save computations. However, this fails to reason over the whole sequence of video, leading to sub-optimal performance. To address this problem, we introduce a state space layer (SSL) into multi-modal Transformer to efficiently integrate global semantics of the video, which mitigates the video information loss caused by frame and region selection modules. Our SSL includes a gating unit to enable controllability over the flow of global semantics into visual representations. To further enhance the controllability, we introduce a cross-modal compositional congruence (C^3) objective to encourage global semantics aligned with the question. To rigorously evaluate long-form videoQA capacity, we construct two new benchmarks Ego-QA and MAD-QA featuring videos of considerably long length, i.e. 17.5 minutes and 1.9 hours, respectively. Extensive experiments demonstrate the superiority of our framework on these new as well as existing datasets.
Authors: Kavosh Asadi, Yao Liu, Shoham Sabach, Ming Yin, Rasool Fakoor
Abstract: We focus on the task of learning the value function in the reinforcement learning (RL) setting. This task is often solved by updating a pair of online and target networks while ensuring that the parameters of these two networks are equivalent. We propose Lookahead-Replicate (LR), a new value-function approximation algorithm that is agnostic to this parameter-space equivalence. Instead, the LR algorithm is designed to maintain an equivalence between the two networks in the function space. This value-based equivalence is obtained by employing a new target-network update. We show that LR leads to a convergent behavior in learning the value function. We also present empirical results demonstrating that LR-based target-network updates significantly improve deep RL on the Atari benchmark.
Authors: Qiyao Ma, Xubin Ren, Chao Huang
Abstract: Recommender systems help users navigate information overload by providing personalized recommendations aligned with their preferences. Collaborative Filtering (CF) is a widely adopted approach, but while advanced techniques like graph neural networks (GNNs) and self-supervised learning (SSL) have enhanced CF models for better user representations, they often lack the ability to provide explanations for the recommended items. Explainable recommendations aim to address this gap by offering transparency and insights into the recommendation decision-making process, enhancing users' understanding. This work leverages the language capabilities of Large Language Models (LLMs) to push the boundaries of explainable recommender systems. We introduce a model-agnostic framework called XRec, which enables LLMs to provide comprehensive explanations for user behaviors in recommender systems. By integrating collaborative signals and designing a lightweight collaborative adaptor, the framework empowers LLMs to understand complex patterns in user-item interactions and gain a deeper understanding of user preferences. Our extensive experiments demonstrate the effectiveness of XRec, showcasing its ability to generate comprehensive and meaningful explanations that outperform baseline approaches in explainable recommender systems. We open-source our model implementation at https://github.com/HKUDS/XRec.
Authors: Sigmund H. H{\o}eg, Yilun Du, Olav Egeland
Abstract: Diffusion models have seen rapid adoption in robotic imitation learning, enabling autonomous execution of complex dexterous tasks. However, action synthesis is often slow, requiring many steps of iterative denoising, limiting the extent to which models can be used in tasks that require fast reactive policies. To sidestep this, recent works have explored how the distillation of the diffusion process can be used to accelerate policy synthesis. However, distillation is computationally expensive and can hurt both the accuracy and diversity of synthesized actions. We propose SDP (Streaming Diffusion Policy), an alternative method to accelerate policy synthesis, leveraging the insight that generating a partially denoised action trajectory is substantially faster than a full output action trajectory. At each observation, our approach outputs a partially denoised action trajectory with variable levels of noise corruption, where the immediate action to execute is noise-free, with subsequent actions having increasing levels of noise and uncertainty. The partially denoised action trajectory for a new observation can then be quickly generated by applying a few steps of denoising to the previously predicted noisy action trajectory (rolled over by one timestep). We illustrate the efficacy of this approach, dramatically speeding up policy synthesis while preserving performance across both simulated and real-world settings.
Authors: Zijia Song, Zelin Zang, Yelin Wang, Guozheng Yang, Kaicheng yu, Wanyu Chen, Miaoyu Wang, Stan Z. Li
Abstract: Multimodal fusion breaks through the boundaries between diverse modalities and has already achieved notable performances. However, in many specialized fields, it is struggling to obtain sufficient alignment data for training, which seriously limits the use of previously effective models. Therefore, semi-supervised learning approaches are attempted to facilitate multimodal alignment by learning from low-alignment data with fewer matched pairs, but traditional techniques like pseudo-labeling may run into troubles in the label-deficient scenarios. To tackle these challenges, we reframe semi-supervised multimodal alignment as a manifold matching issue and propose a new methodology based on CLIP, termed Set-CLIP. Specifically, by designing a novel semantic density distribution loss, we constrain the latent representation distribution with fine granularity and extract implicit semantic alignment from unpaired multimodal data, thereby reducing the reliance on numerous strictly matched pairs. Furthermore, we apply coarse-grained modality adaptation and unimodal self-supervised guidance to narrow the gaps between modality spaces and improve the stability of representation distributions. Extensive experiments conducted on a range of tasks in various fields, including protein analysis, remote sensing, and the general vision-language field, validate the efficacy of our proposed Set-CLIP method. Especially with no paired data for supervised training, Set-CLIP is still outstanding, which brings an improvement of 144.83% over CLIP.
Authors: Jinqiang Wang, Huansheng Ning, Yi Peng, Qikai Wei, Daniel Tesfai, Wenwei Mao, Tao Zhu, Runhe Huang
Abstract: Large Language Models (LLMs) have demonstrated surprising performance across various natural language processing tasks. Recently, medical LLMs enhanced with domain-specific knowledge have exhibited excellent capabilities in medical consultation and diagnosis. These models can smoothly simulate doctor-patient dialogues and provide professional medical advice. Most medical LLMs are developed through continued training of open-source general LLMs, which require significantly fewer computational resources than training LLMs from scratch. Additionally, this approach offers better patient privacy protection than API-based solutions. Given the above advantages, this survey systematically summarizes how to train medical LLMs based on open-source general LLMs from a more fine-grained perspective. It covers (a) how to acquire training corpus and construct customized medical training sets, (b) how to choose an appropriate training paradigm, (c) how to choose a suitable evaluation benchmark, and (d) existing challenges and promising research directions are discussed. This survey can provide guidance for the development of LLMs focused on various medical applications, such as medical education, diagnostic planning, and clinical assistants. Related resources and supplemental information can be found on the GitHub repository.
Authors: Alessio Devoto, Yu Zhao, Simone Scardapane, Pasquale Minervini
Abstract: The deployment of large language models (LLMs) is often hindered by the extensive memory requirements of the Key-Value (KV) cache, especially as context lengths increase. Existing approaches to reduce the KV cache size involve either fine-tuning the model to learn a compression strategy or leveraging attention scores to reduce the sequence length. We analyse the attention distributions in decoder-only Transformers-based models and observe that attention allocation patterns stay consistent across most layers. Surprisingly, we find a clear correlation between the $L_2$ and the attention scores over cached KV pairs, where a low $L_2$ of a key embedding usually leads to a high attention score during decoding. This finding indicates that the influence of a KV pair is potentially determined by the key embedding itself before being queried. Based on this observation, we compress the KV cache based on the $L_2$ of key embeddings. Our experimental results show that this simple strategy can reduce the KV cache size by 50% on language modelling and needle-in-a-haystack tasks and 90% on passkey retrieval tasks without losing accuracy. Moreover, without relying on the attention scores, this approach remains compatible with FlashAttention, enabling broader applicability.
Authors: Sherzod Hakimov, Yerkezhan Abdullayeva, Kushal Koshti, Antonia Schmidt, Yan Weiser, Anne Beyer, David Schlangen
Abstract: While the situation has improved for text-only models, it again seems to be the case currently that multimodal (text and image) models develop faster than ways to evaluate them. In this paper, we bring a recently developed evaluation paradigm from text models to multimodal models, namely evaluation through the goal-oriented game (self) play, complementing reference-based and preference-based evaluation. Specifically, we define games that challenge a model's capability to represent a situation from visual information and align such representations through dialogue. We find that the largest closed models perform rather well on the games that we define, while even the best open-weight models struggle with them. On further analysis, we find that the exceptional deep captioning capabilities of the largest models drive some of the performance. There is still room to grow for both kinds of models, ensuring the continued relevance of the benchmark.
Authors: Zhiyuan Chang, Mingyang Li, Junjie Wang, Yi Liu, Qing Wang, Yang Liu
Abstract: Text-to-Image Diffusion Models (T2I DMs) have garnered significant attention for their ability to generate high-quality images from textual descriptions. However, these models often produce images that do not fully align with the input prompts, resulting in semantic inconsistencies. The most prominent issue among these semantic inconsistencies is catastrophic-neglect, where the images generated by T2I DMs miss key objects mentioned in the prompt. We first conduct an empirical study on this issue, exploring the prevalence of catastrophic-neglect, potential mitigation strategies with feature enhancement, and the insights gained. Guided by the empirical findings, we propose an automated repair approach named Patcher to address catastrophic-neglect in T2I DMs. Specifically, Patcher first determines whether there are any neglected objects in the prompt, and then applies attention-guided feature enhancement to these neglected objects, resulting in a repaired prompt. Experimental results on three versions of Stable Diffusion demonstrate that Patcher effectively repairs the issue of catastrophic-neglect, achieving 10.1%-16.3% higher Correct Rate in image generation compared to baselines.
Authors: Shayan Kiyani, George Pappas, Hamed Hassani
Abstract: Conditional validity and length efficiency are two crucial aspects of conformal prediction (CP). Achieving conditional validity ensures accurate uncertainty quantification for data subpopulations, while proper length efficiency ensures that the prediction sets remain informative and non-trivial. Despite significant efforts to address each of these issues individually, a principled framework that reconciles these two objectives has been missing in the CP literature. In this paper, we develop Conformal Prediction with Length-Optimization (CPL) - a novel framework that constructs prediction sets with (near-) optimal length while ensuring conditional validity under various classes of covariate shifts, including the key cases of marginal and group-conditional coverage. In the infinite sample regime, we provide strong duality results which indicate that CPL achieves conditional validity and length optimality. In the finite sample regime, we show that CPL constructs conditionally valid prediction sets. Our extensive empirical evaluations demonstrate the superior prediction set size performance of CPL compared to state-of-the-art methods across diverse real-world and synthetic datasets in classification, regression, and large language model-based multiple choice question answering.
Authors: Shengcheng Luo, Quanquan Peng, Jun Lv, Kaiwen Hong, Katherine Rose Driggs-Campbell, Cewu Lu, Yong-Lu Li
Abstract: Employing a teleoperation system for gathering demonstrations offers the potential for more efficient learning of robot manipulation. However, teleoperating a robot arm equipped with a dexterous hand or gripper, via a teleoperation system presents inherent challenges due to the task's high dimensionality, complexity of motion, and differences between physiological structures. In this study, we introduce a novel system for joint learning between human operators and robots, that enables human operators to share control of a robot end-effector with a learned assistive agent, simplifies the data collection process, and facilitates simultaneous human demonstration collection and robot manipulation training. As data accumulates, the assistive agent gradually learns. Consequently, less human effort and attention are required, enhancing the efficiency of the data collection process. It also allows the human operator to adjust the control ratio to achieve a trade-off between manual and automated control. We conducted experiments in both simulated environments and physical real-world settings. Through user studies and quantitative evaluations, it is evident that the proposed system could enhance data collection efficiency and reduce the need for human adaptation while ensuring the collected data is of sufficient quality for downstream tasks. \textit{For more details, please refer to our webpage https://norweig1an.github.io/HAJL.github.io/.
Authors: Noam Koren, Abigail Goldsteen, Guy Amit, Ariel Farkash
Abstract: Analyzing time-series data that contains personal information, particularly in the medical field, presents serious privacy concerns. Sensitive health data from patients is often used to train machine learning models for diagnostics and ongoing care. Assessing the privacy risk of such models is crucial to making knowledgeable decisions on whether to use a model in production or share it with third parties. Membership Inference Attacks (MIA) are a key method for this kind of evaluation, however time-series prediction models have not been thoroughly studied in this context. We explore existing MIA techniques on time-series models, and introduce new features, focusing on the seasonality and trend components of the data. Seasonality is estimated using a multivariate Fourier transform, and a low-degree polynomial is used to approximate trends. We applied these techniques to various types of time-series models, using datasets from the health domain. Our results demonstrate that these new features enhance the effectiveness of MIAs in identifying membership, improving the understanding of privacy risks in medical data applications.
Authors: Rachael Fleurence, Jiang Bian, Xiaoyan Wang, Hua Xu, Dalia Dawoud, Mitch Higashi, Jagpreet Chhatwal
Abstract: This review introduces the transformative potential of generative Artificial Intelligence (AI) and foundation models, including large language models (LLMs), for health technology assessment (HTA). We explore their applications in four critical areas, evidence synthesis, evidence generation, clinical trials and economic modeling: (1) Evidence synthesis: Generative AI has the potential to assist in automating literature reviews and meta-analyses by proposing search terms, screening abstracts, and extracting data with notable accuracy; (2) Evidence generation: These models can potentially facilitate automating the process and analyze the increasingly available large collections of real-world data (RWD), including unstructured clinical notes and imaging, enhancing the speed and quality of real-world evidence (RWE) generation; (3) Clinical trials: Generative AI can be used to optimize trial design, improve patient matching, and manage trial data more efficiently; and (4) Economic modeling: Generative AI can also aid in the development of health economic models, from conceptualization to validation, thus streamlining the overall HTA process. Despite their promise, these technologies, while rapidly improving, are still nascent and continued careful evaluation in their applications to HTA is required. To ensure their responsible use and implementation, both developers and users of research incorporating these tools, should familiarize themselves with their current limitations, including the issues related to scientific validity, risk of bias, and consider equity and ethical implications. We also surveyed the current policy landscape and provide suggestions for HTA agencies on responsibly integrating generative AI into their workflows, emphasizing the importance of human oversight and the fast-evolving nature of these tools.
Authors: Jos\'e Daniel Pascual-Triana, Alberto Fern\'andez, Paulo Novais, Francisco Herrera
Abstract: One of the key issues regarding classification problems in Trustworthy Artificial Intelligence is ensuring Fairness in the prediction of different classes when protected (sensitive) features are present. Data quality is critical in these cases, as biases in training data can be reflected in machine learning, impacting human lives and failing to comply with current regulations. One strategy to improve data quality and avoid these problems is preprocessing the dataset. Instance selection via undersampling can foster balanced learning of classes and protected feature values. Performing undersampling in class overlap areas close to the decision boundary should bolster the impact on the classifier. This work proposes Fair Overlap Number of Balls (Fair-ONB), an undersampling method that harnesses the data morphology of the different data groups (obtained from the combination of classes and protected feature values) to perform guided undersampling in overlap areas. It employs attributes of the ball coverage of the groups, such as the radius, number of covered instances and density, to select the most suitable areas for undersampling and reduce bias. Results show that the Fair-ONB method improves model Fairness with low impact on the classifier's predictive performance.
Authors: Lorenzo Olearo, Giorgio Longari, Simone Melzi, Alessandro Raganato, Rafael Pe\~naloza
Abstract: For the last decade, there has been a push to use multi-dimensional (latent) spaces to represent concepts; and yet how to manipulate these concepts or reason with them remains largely unclear. Some recent methods exploit multiple latent representations and their connection, making this research question even more entangled. Our goal is to understand how operations in the latent space affect the underlying concepts. To that end, we explore the task of concept blending through diffusion models. Diffusion models are based on a connection between a latent representation of textual prompts and a latent space that enables image reconstruction and generation. This task allows us to try different text-based combination strategies, and evaluate easily through a visual analysis. Our conclusion is that concept blending through space manipulation is possible, although the best strategy depends on the context of the blend.
Authors: Yilie Huang, Yanwei Jia, Xun Yu Zhou
Abstract: We study reinforcement learning (RL) for a class of continuous-time linear-quadratic (LQ) control problems for diffusions, where states are scalar-valued and running control rewards are absent but volatilities of the state processes depend on both state and control variables. We apply a model-free approach that relies neither on knowledge of model parameters nor on their estimations, and devise an actor-critic algorithm to learn the optimal policy parameter directly. Our main contributions include the introduction of an exploration schedule and a regret analysis of the proposed algorithm. We provide the convergence rate of the policy parameter to the optimal one, and prove that the algorithm achieves a regret bound of $O(N^{\frac{3}{4}})$ up to a logarithmic factor, where $N$ is the number of learning episodes. We conduct a simulation study to validate the theoretical results and demonstrate the effectiveness and reliability of the proposed algorithm. We also perform numerical comparisons between our method and those of the recent model-based stochastic LQ RL studies adapted to the state- and control-dependent volatility setting, demonstrating a better performance of the former in terms of regret bounds.
Authors: Chong Zhang, Xinyi Liu, Zhongmou Zhang, Mingyu Jin, Lingyao Li, Zhenting Wang, Wenyue Hua, Dong Shu, Suiyuan Zhu, Xiaobo Jin, Sujian Li, Mengnan Du, Yongfeng Zhang
Abstract: Can AI Agents simulate real-world trading environments to investigate the impact of external factors on stock trading activities (e.g., macroeconomics, policy changes, company fundamentals, and global events)? These factors, which frequently influence trading behaviors, are critical elements in the quest for maximizing investors' profits. Our work attempts to solve this problem through large language model based agents. We have developed a multi-agent AI system called StockAgent, driven by LLMs, designed to simulate investors' trading behaviors in response to the real stock market. The StockAgent allows users to evaluate the impact of different external factors on investor trading and to analyze trading behavior and profitability effects. Additionally, StockAgent avoids the test set leakage issue present in existing trading simulation systems based on AI Agents. Specifically, it prevents the model from leveraging prior knowledge it may have acquired related to the test data. We evaluate different LLMs under the framework of StockAgent in a stock trading environment that closely resembles real-world conditions. The experimental results demonstrate the impact of key external factors on stock market trading, including trading behavior and stock price fluctuation rules. This research explores the study of agents' free trading gaps in the context of no prior knowledge related to market data. The patterns identified through StockAgent simulations provide valuable insights for LLM-based investment advice and stock recommendation. The code is available at https://github.com/MingyuJ666/Stockagent.
Authors: Niklas Wretblad, Oskar Holmstr\"om, Erik Larsson, Axel Wiks\"ater, Oscar S\"oderlund, Hjalmar \"Ohman, Ture Pont\'en, Martin Forsberg, Martin S\"orme, Fredrik Heintz
Abstract: Relational databases often suffer from uninformative descriptors of table contents, such as ambiguous columns and hard-to-interpret values, impacting both human users and text-to-SQL models. In this paper, we explore the use of large language models (LLMs) to automatically generate detailed natural language descriptions for SQL database columns, aiming to improve text-to-SQL performance and automate metadata creation. We create a dataset of gold column descriptions based on the BIRD-Bench benchmark, manually refining its column descriptions and creating a taxonomy for categorizing column difficulty. Through evaluating several LLMs, we find that incorporating these column descriptions consistently enhances text-to-SQL model performance, particularly for larger models like GPT-4o, Qwen2 72B and Mixtral 22Bx8. However, models struggle with columns that exhibit inherent ambiguity, highlighting the need for manual expert input. Notably, Qwen2-generated descriptions, containing by annotators deemed superfluous information, outperform manually curated gold descriptions, suggesting that models benefit from more detailed metadata than humans expect. Future work will investigate the specific features of these high-performing descriptions and explore other types of metadata, such as numerical reasoning and synonyms, to further improve text-to-SQL systems. The dataset, annotations and code will all be made available.
Authors: Zhibo Zhang, Wuxia Bai, Yuxi Li, Mark Huasong Meng, Kailong Wang, Ling Shi, Li Li, Jun Wang, Haoyu Wang
Abstract: Large language models (LLMs) have achieved unprecedented success in the field of natural language processing. However, the black-box nature of their internal mechanisms has brought many concerns about their trustworthiness and interpretability. Recent research has discovered a class of abnormal tokens in the model's vocabulary space and named them "glitch tokens". Those tokens, once included in the input, may induce the model to produce incorrect, irrelevant, or even harmful results, drastically undermining the reliability and practicality of LLMs. In this work, we aim to enhance the understanding of glitch tokens and propose techniques for their detection and mitigation. We first reveal the characteristic features induced by glitch tokens on LLMs, which are evidenced by significant deviations in the distributions of attention patterns and dynamic information from intermediate model layers. Based on the insights, we develop GlitchProber, a tool for efficient glitch token detection and mitigation. GlitchProber utilizes small-scale sampling, principal component analysis for accelerated feature extraction, and a simple classifier for efficient vocabulary screening. Taking one step further, GlitchProber rectifies abnormal model intermediate layer values to mitigate the destructive effects of glitch tokens. Evaluated on five mainstream open-source LLMs, GlitchProber demonstrates higher efficiency, precision, and recall compared to existing approaches, with an average F1 score of 0.86 and an average repair rate of 50.06%. GlitchProber unveils a novel path to address the challenges posed by glitch tokens and inspires future research toward more robust and interpretable LLMs.
Authors: Julian Rodemann, Christoph Jansen, Georg Schollmeyer
Abstract: We demonstrate that a wide array of machine learning algorithms are specific instances of one single paradigm: reciprocal learning. These instances range from active learning over multi-armed bandits to self-training. We show that all these algorithms do not only learn parameters from data but also vice versa: They iteratively alter training data in a way that depends on the current model fit. We introduce reciprocal learning as a generalization of these algorithms using the language of decision theory. This allows us to study under what conditions they converge. The key is to guarantee that reciprocal learning contracts such that the Banach fixed-point theorem applies. In this way, we find that reciprocal learning algorithms converge at linear rates to an approximately optimal model under relatively mild assumptions on the loss function, if their predictions are probabilistic and the sample adaption is both non-greedy and either randomized or regularized. We interpret these findings and provide corollaries that relate them to specific active learning, self-training, and bandit algorithms.
Authors: Anders Gj{\o}lbye, Lina Skerath, William Lehn-Schi{\o}ler, Nicolas Langer, Lars Kai Hansen
Abstract: Electroencephalography (EEG) research typically focuses on tasks with narrowly defined objectives, but recent studies are expanding into the use of unlabeled data within larger models, aiming for a broader range of applications. This addresses a critical challenge in EEG research. For example, Kostas et al. (2021) show that self-supervised learning (SSL) outperforms traditional supervised methods. Given the high noise levels in EEG data, we argue that further improvements are possible with additional preprocessing. Current preprocessing methods often fail to efficiently manage the large data volumes required for SSL, due to their lack of optimization, reliance on subjective manual corrections, and validation processes or inflexible protocols that limit SSL. We propose a Python-based EEG preprocessing pipeline optimized for self-supervised learning, designed to efficiently process large-scale data. This optimization not only stabilizes self-supervised training but also enhances performance on downstream tasks compared to training with raw data.
Authors: Jiuang Zhao, Donghao Yang, Li Zhang, Xiaoli Lian, Zitian Yang, Fang Liu
Abstract: Automatic Program Repair (APR) endeavors to autonomously rectify issues within specific projects, which generally encompasses three categories of tasks: bug resolution, new feature development, and feature enhancement. Despite extensive research proposing various methodologies, their efficacy in addressing real issues remains unsatisfactory. It's worth noting that, typically, engineers have design rationales (DR) on solution-planed solutions and a set of underlying reasons-before they start patching code. In open-source projects, these DRs are frequently captured in issue logs through project management tools like Jira. This raises a compelling question: How can we leverage DR scattered across the issue logs to efficiently enhance APR? To investigate this premise, we introduce DRCodePilot, an approach designed to augment GPT-4-Turbo's APR capabilities by incorporating DR into the prompt instruction. Furthermore, given GPT-4's constraints in fully grasping the broader project context and occasional shortcomings in generating precise identifiers, we have devised a feedback-based self-reflective framework, in which we prompt GPT-4 to reconsider and refine its outputs by referencing a provided patch and suggested identifiers. We have established a benchmark comprising 938 issue-patch pairs sourced from two open-source repositories hosted on GitHub and Jira. Our experimental results are impressive: DRCodePilot achieves a full-match ratio that is a remarkable 4.7x higher than when GPT-4 is utilized directly. Additionally, the CodeBLEU scores also exhibit promising enhancements. Moreover, our findings reveal that the standalone application of DR can yield promising increase in the full-match ratio across CodeLlama, GPT-3.5, and GPT-4 within our benchmark suite. We believe that our DRCodePilot initiative heralds a novel human-in-the-loop avenue for advancing the field of APR.
Authors: Purushothaman Natarajan, Athira Nambiar
Abstract: Deep learning techniques have revolutionized image classification by mimicking human cognition and automating complex decision-making processes. However, the deployment of AI systems in the wild, especially in high-security domains such as defence, is curbed by the lack of explainability of the model. To this end, eXplainable AI (XAI) is an emerging area of research that is intended to explore the unexplained hidden black box nature of deep neural networks. This paper explores the application of the eXplainable Artificial Intelligence (XAI) tool to interpret the underwater image classification results, one of the first works in the domain to the best of our knowledge. Our study delves into the realm of SONAR image classification using a custom dataset derived from diverse sources, including the Seabed Objects KLSG dataset, the camera SONAR dataset, the mine SONAR images dataset, and the SCTD dataset. An extensive analysis of transfer learning techniques for image classification using benchmark Convolutional Neural Network (CNN) architectures such as VGG16, ResNet50, InceptionV3, DenseNet121, etc. is carried out. On top of this classification model, a post-hoc XAI technique, viz. Local Interpretable Model-Agnostic Explanations (LIME) are incorporated to provide transparent justifications for the model's decisions by perturbing input data locally to see how predictions change. Furthermore, Submodular Picks LIME (SP-LIME) a version of LIME particular to images, that perturbs the image based on the submodular picks is also extensively studied. To this end, two submodular optimization algorithms i.e. Quickshift and Simple Linear Iterative Clustering (SLIC) are leveraged towards submodular picks. The extensive analysis of XAI techniques highlights interpretability of the results in a more human-compliant way, thus boosting our confidence and reliability.
Authors: Xiangxiang Shen, Zheng Wan, Lingfeng Wen, Licheng Sun, Ou Yang Ming Jie, Xuan Tang, Xian Zeng, Mingsong Chen, Xiao He, Xian Wei
Abstract: To date, the International Zeolite Association Structure Commission (IZA-SC) has cataloged merely 255 distinct zeolite structures, with millions of theoretically possible structures yet to be discovered. The synthesis of a specific zeolite typically necessitates the use of an organic structure-directing agent (OSDA), since the selectivity for a particular zeolite is largely determined by the affinity between the OSDA and the zeolite. Therefore, finding the best affinity OSDA-zeolite pair is the key to the synthesis of targeted zeolite. However, OSDA-zeolite pairs frequently exhibit complex geometric structures, i.e., a complex crystal structure formed by a large number of atoms. Although some existing machine learning methods can represent the periodicity of crystals, they cannot accurately represent crystal structures with local variability. To address this issue, we propose a novel approach called Zeoformer, which can effectively represent coarse-grained crystal periodicity and fine-grained local variability. Zeoformer reconstructs the unit cell centered around each atom and encodes the pairwise distances between this central atom and other atoms within the reconstructed unit cell. The introduction of pairwise distances within the reconstructed unit cell more effectively represents the overall structure of the unit cell and the differences between different unit cells, enabling the model to more accurately and efficiently predict the properties of OSDA-zeolite pairs and general crystal structures. Through comprehensive evaluation, our Zeoformer model demonstrates the best performance on OSDA-zeolite pair datasets and two types of crystal material datasets.
Authors: Ben Batten, Yang Zheng, Alessandro De Palma, Panagiotis Kouvaros, Alessio Lomuscio
Abstract: We address the problem of verifying neural networks against geometric transformations of the input image, including rotation, scaling, shearing, and translation. The proposed method computes provably sound piecewise linear constraints for the pixel values by using sampling and linear approximations in combination with branch-and-bound Lipschitz optimisation. The method obtains provably tighter over-approximations of the perturbation region than the present state-of-the-art. We report results from experiments on a comprehensive set of verification benchmarks on MNIST and CIFAR10. We show that our proposed implementation resolves up to 32% more verification cases than present approaches.
Authors: Hanna Abi Akl
Abstract: We introduce SHADOW, a fine-tuned language model trained on an intermediate task using associative deductive reasoning, and measure its performance on a knowledge base construction task using Wikidata triple completion. We evaluate SHADOW on the LM-KBC 2024 challenge and show that it outperforms the baseline solution by 20% with a F1 score of 68.72%.
Authors: Yang Liu, Xichou Zhu, Zhou Shen, Yi Liu, Min Li, Yujun Chen, Benzi John, Zhenzhen Ma, Tao Hu, Zhi Li, Zhiyang Xu, Wei Luo, Junhui Wang
Abstract: Large Language Models (LLMs) have recently displayed their extraordinary capabilities in language understanding. However, how to comprehensively assess the sentiment capabilities of LLMs continues to be a challenge. This paper investigates the ability of LLMs to detect and react to sentiment in text modal. As the integration of LLMs into diverse applications is on the rise, it becomes highly critical to comprehend their sensitivity to emotional tone, as it can influence the user experience and the efficacy of sentiment-driven tasks. We conduct a series of experiments to evaluate the performance of several prominent LLMs in identifying and responding appropriately to sentiments like positive, negative, and neutral emotions. The models' outputs are analyzed across various sentiment benchmarks, and their responses are compared with human evaluations. Our discoveries indicate that although LLMs show a basic sensitivity to sentiment, there are substantial variations in their accuracy and consistency, emphasizing the requirement for further enhancements in their training processes to better capture subtle emotional cues. Take an example in our findings, in some cases, the models might wrongly classify a strongly positive sentiment as neutral, or fail to recognize sarcasm or irony in the text. Such misclassifications highlight the complexity of sentiment analysis and the areas where the models need to be refined. Another aspect is that different LLMs might perform differently on the same set of data, depending on their architecture and training datasets. This variance calls for a more in-depth study of the factors that contribute to the performance differences and how they can be optimized.
Authors: Guanwen Xie, Jingzehua Xu, Yiyuan Yang, Yimian Ding, Shuai Zhang
Abstract: Achieving the effective design and improvement of reward functions in reinforcement learning (RL) tasks with complex custom environments and multiple requirements presents considerable challenges. In this paper, we propose ERFSL, an efficient reward function searcher using LLMs, which enables LLMs to be effective white-box searchers and highlights their advanced semantic understanding capabilities. Specifically, we generate reward components for each numerically explicit user requirement and employ a reward critic to identify the correct code form. Then, LLMs assign weights to the reward components to balance their values and iteratively adjust the weights without ambiguity and redundant adjustments by flexibly adopting directional mutation and crossover strategies, similar to genetic algorithms, based on the context provided by the training log analyzer. We applied the framework to an underwater data collection RL task without direct human feedback or reward examples (zero-shot learning). The reward critic successfully corrects the reward code with only one feedback instance for each requirement, effectively preventing unrectifiable errors. The initialization of weights enables the acquisition of different reward functions within the Pareto solution set without the need for weight search. Even in cases where a weight is 500 times off, on average, only 5.2 iterations are needed to meet user requirements. The ERFSL also works well with most prompts utilizing GPT-4o mini, as we decompose the weight searching process to reduce the requirement for numerical and long-context understanding capabilities
Authors: Kaiwen Zheng, Yongxin Chen, Hanzi Mao, Ming-Yu Liu, Jun Zhu, Qinsheng Zhang
Abstract: Masked diffusion models (MDMs) have emerged as a popular research topic for generative modeling of discrete data, thanks to their superior performance over other discrete diffusion models, and are rivaling the auto-regressive models (ARMs) for language modeling tasks. The recent effort in simplifying the masked diffusion framework further leads to alignment with continuous-space diffusion models and more principled training and sampling recipes. In this paper, however, we reveal that both training and sampling of MDMs are theoretically free from the time variable, arguably the key signature of diffusion models, and are instead equivalent to masked models. The connection on the sampling aspect is drawn by our proposed first-hitting sampler (FHS). Specifically, we show that the FHS is theoretically equivalent to MDMs' original generation process while significantly alleviating the time-consuming categorical sampling and achieving a 20$\times$ speedup. In addition, our investigation raises doubts about whether MDMs can truly beat ARMs. We identify, for the first time, an underlying numerical issue, even with the commonly used 32-bit floating-point precision, which results in inaccurate categorical sampling. We show that the numerical issue lowers the effective temperature both theoretically and empirically, and the resulting decrease in token diversity makes previous evaluations, which assess the generation quality solely through the incomplete generative perplexity metric, somewhat unfair.
Authors: Zhuowei Chen, Lianxi Wang, Yuben Wu, Xinfeng Liao, Yujia Tian, Junyang Zhong
Abstract: Sentiment classification (SC) often suffers from low-resource challenges such as domain-specific contexts, imbalanced label distributions, and few-shot scenarios. The potential of the diffusion language model (LM) for textual data augmentation (DA) remains unexplored, moreover, textual DA methods struggle to balance the diversity and consistency of new samples. Most DA methods either perform logical modifications or rephrase less important tokens in the original sequence with the language model. In the context of SC, strong emotional tokens could act critically on the sentiment of the whole sequence. Therefore, contrary to rephrasing less important context, we propose DiffusionCLS to leverage a diffusion LM to capture in-domain knowledge and generate pseudo samples by reconstructing strong label-related tokens. This approach ensures a balance between consistency and diversity, avoiding the introduction of noise and augmenting crucial features of datasets. DiffusionCLS also comprises a Noise-Resistant Training objective to help the model generalize. Experiments demonstrate the effectiveness of our method in various low-resource scenarios including domain-specific and domain-general problems. Ablation studies confirm the effectiveness of our framework's modules, and visualization studies highlight optimal deployment conditions, reinforcing our conclusions.
Authors: Wei Wen, Quanyu Zhu, Weiwei Chu, Wen-Yen Chen, Jiyan Yang
Abstract: Scaling up deep learning models has been proven effective to improve intelligence of machine learning (ML) models, especially for industry recommendation models and large language models. The co-design of large distributed ML systems and algorithms (to maximize training performance) plays a pivotal role for its success. As it scales, the number of co-design hyper-parameters grows rapidly which brings challenges to feasibly find the optimal setup for system performance maximization. In this paper, we propose CubicML which uses ML to automatically optimize training performance of large distributed ML systems. In CubicML, we use an ML model as a proxy to predict the training performance for search efficiency and performance modeling flexibility. We proved that CubicML can effectively optimize training speed of in-house ads recommendation models with 73 billion parameters and large language models up to 405 billion parameters at Meta.
Authors: Wenyu Zhang, Shuo Sun, Bin Wang, Xunlong Zou, Zhuohan Liu, Yingxu He, Geyu Lin, Nancy F. Chen, Ai Ti Aw
Abstract: The rapid advancements in large language models (LLMs) have significantly enhanced natural language processing capabilities, facilitating the development of AudioLLMs that process and understand speech and audio inputs alongside text. Existing AudioLLMs typically combine a pre-trained audio encoder with a pre-trained LLM, which are subsequently finetuned on specific audio tasks. However, the pre-trained audio encoder has constrained capacity to capture features for new tasks and datasets. To address this, we propose to incorporate mixtures of `weak' encoders (MoWE) into the AudioLLM framework. MoWE supplements a base encoder with a pool of relatively light weight encoders, selectively activated based on the audio input to enhance feature extraction without significantly increasing model size. Our empirical results demonstrate that MoWE effectively improves multi-task performance, broadening the applicability of AudioLLMs to more diverse audio tasks.
Authors: Guiliang Liu, Sheng Xu, Shicheng Liu, Ashish Gaurav, Sriram Ganapathi Subramanian, Pascal Poupart
Abstract: Inverse Constrained Reinforcement Learning (ICRL) is the task of inferring the implicit constraints followed by expert agents from their demonstration data. As an emerging research topic, ICRL has received considerable attention in recent years. This article presents a categorical survey of the latest advances in ICRL. It serves as a comprehensive reference for machine learning researchers and practitioners, as well as starters seeking to comprehend the definitions, advancements, and important challenges in ICRL. We begin by formally defining the problem and outlining the algorithmic framework that facilitates constraint inference across various scenarios. These include deterministic or stochastic environments, environments with limited demonstrations, and multiple agents. For each context, we illustrate the critical challenges and introduce a series of fundamental methods to tackle these issues. This survey encompasses discrete, virtual, and realistic environments for evaluating ICRL agents. We also delve into the most pertinent applications of ICRL, such as autonomous driving, robot control, and sports analytics. To stimulate continuing research, we conclude the survey with a discussion of key unresolved questions in ICRL that can effectively foster a bridge between theoretical understanding and practical industrial applications.
Authors: Yufei Tao, Adam Hiatt, Erik Haake, Antonie J. Jetter, Ameeta Agrawal
Abstract: Large language models (LLMs) have demonstrated remarkable progress in leveraging diverse knowledge sources. This study investigates how nine widely used LLMs allocate knowledge between local context and global parameters when answering open-ended questions in knowledge-consistent scenarios. We introduce a novel dataset, WikiAtomic, and systematically vary context sizes to analyze how LLMs prioritize and utilize the provided information and their parametric knowledge in knowledge-consistent scenarios. Additionally, we also study their tendency to hallucinate under varying context sizes. Our findings reveal consistent patterns across models, including a consistent reliance on both contextual (around 70%) and parametric (around 30%) knowledge, and a decrease in hallucinations with increasing context. These insights highlight the importance of more effective context organization and developing models that use input more deterministically for robust performance.
Authors: Yanlin Wang, Wanjun Zhong, Yanxian Huang, Ensheng Shi, Min Yang, Jiachi Chen, Hui Li, Yuchi Ma, Qianxiang Wang, Zibin Zheng
Abstract: In recent years, Large Language Models (LLMs) have achieved remarkable success and have been widely used in various downstream tasks, especially in the tasks of the software engineering (SE) field. We find that many studies combining LLMs with SE have employed the concept of agents either explicitly or implicitly. However, there is a lack of an in-depth survey to sort out the development context of existing works, analyze how existing works combine the LLM-based agent technologies to optimize various tasks, and clarify the framework of LLM-based agents in SE. In this paper, we conduct the first survey of the studies on combining LLM-based agents with SE and present a framework of LLM-based agents in SE which includes three key modules: perception, memory, and action. We also summarize the current challenges in combining the two fields and propose future opportunities in response to existing challenges. We maintain a GitHub repository of the related papers at: https://github.com/DeepSoftwareAnalytics/Awesome-Agent4SE.
URLs: https://github.com/DeepSoftwareAnalytics/Awesome-Agent4SE.
Authors: Xiang Zhang, Muhammad Abdul-Mageed, Laks V. S. Lakshmanan
Abstract: The Transformer architecture excels in a variety of language modeling tasks, outperforming traditional neural architectures such as RNN and LSTM. This is partially due to its elimination of recurrent connections, which allows for parallel training and a smoother flow of gradients. However, this move away from recurrent structures places the Transformer model at the lower end of Chomsky's computational hierarchy, imposing limitations on its computational abilities. Consequently, even advanced Transformer-based models face considerable difficulties in tasks like counting, string reversal, and multiplication. These tasks, though seemingly elementary, require a level of computational complexity that exceeds the capabilities of the Transformer architecture. Concurrently, the emergence of ``Chain of Thought" (CoT) prompting has enabled Transformer-based language models to tackle tasks that were previously impossible or poorly executed. In this work, we thoroughly investigate the influence of recurrent structures in neural models on their reasoning abilities and computability, contrasting the role autoregression plays in the neural models' computational power. We then shed light on how the CoT approach can mimic recurrent computation and act as a bridge between autoregression and recurrence in the context of language models. It is this approximated recurrence that notably improves the model's performance and computational capacity. Moreover, we revisit recent recurrent-based Transformer model designs, focusing on their computational abilities through our proposed concept of ``recurrence-completeness" and identify key theoretical limitations in models like Linear Transformer and RWKV. Through this, we aim to provide insight into the neural model architectures and prompt better model design.
Authors: Yi-Jyun Sun, Suvodip Dey, Dilek Hakkani-Tur, Gokhan Tur
Abstract: Estimation of a model's confidence on its outputs is critical for Conversational AI systems based on large language models (LLMs), especially for reducing hallucination and preventing over-reliance. In this work, we provide an exhaustive exploration of methods, including approaches proposed for open- and closed-weight LLMs, aimed at quantifying and leveraging model uncertainty to improve the reliability of LLM-generated responses, specifically focusing on dialogue state tracking (DST) in task-oriented dialogue systems (TODS). Regardless of the model type, well-calibrated confidence scores are essential to handle uncertainties, thereby improving model performance. We evaluate four methods for estimating confidence scores based on softmax, raw token scores, verbalized confidences, and a combination of these methods, using the area under the curve (AUC) metric to assess calibration, with higher AUC indicating better calibration. We also enhance these with a self-probing mechanism, proposed for closed models. Furthermore, we assess these methods using an open-weight model fine-tuned for the task of DST, achieving superior joint goal accuracy (JGA). Our findings also suggest that fine-tuning open-weight LLMs can result in enhanced AUC performance, indicating better confidence score calibration.
Authors: Shanshan Wang, Chun Zhang, Ning Zhang
Abstract: The Knowledge Graph-to-Text Generation task aims to convert structured knowledge graphs into coherent and human-readable natural language text. Recent efforts in this field have focused on enhancing pre-trained language models (PLMs) by incorporating graph structure information to capture the intricate structure details of knowledge graphs. However, most of these approaches tend to capture only single-granularity structure information, concentrating either on the relationships between entities within the original graph or on the relationships between words within the same entity or across different entities. This narrow focus results in a significant limitation: models that concentrate solely on entity-level structure fail to capture the nuanced semantic relationships between words, while those that focus only on word-level structure overlook the broader relationships between original entire entities. To overcome these limitations, this paper introduces the Multi-granularity Graph Structure Attention (MGSA), which is based on PLMs. The encoder of the model architecture features an entity-level structure encoding module, a word-level structure encoding module, and an aggregation module that synthesizes information from both structure. This multi-granularity structure encoding approach allows the model to simultaneously capture both entity-level and word-level structure information, providing a more comprehensive understanding of the knowledge graph's structure information, thereby significantly improving the quality of the generated text. We conducted extensive evaluations of the MGSA model using two widely recognized KG-to-Text Generation benchmark datasets, WebNLG and EventNarrative, where it consistently outperformed models that rely solely on single-granularity structure information, demonstrating the effectiveness of our approach.
Authors: Y. Isabel Liu, Windsor Nguyen, Yagiz Devre, Evan Dogariu, Anirudha Majumdar, Elad Hazan
Abstract: This paper describes an efficient, open source PyTorch implementation of the Spectral Transform Unit. We investigate sequence prediction tasks over several modalities including language, robotics, and simulated dynamical systems. We find that for the same parameter count, the STU and its variants outperform the Transformer as well as other leading state space models across various modalities.
Authors: Luning Wang, Shiyao Li, Xuefei Ning, Zhihang Yuan, Shengen Yan, Guohao Dai, Yu Wang
Abstract: Large Language Models (LLMs) have been widely adopted to process long-context tasks. However, the large memory overhead of the key-value (KV) cache poses significant challenges in long-context scenarios. Existing training-free KV cache compression methods typically focus on quantization and token pruning, which have compression limits, and excessive sparsity can lead to severe performance degradation. Other methods design new architectures with less KV overhead but require significant training overhead. To address the above two drawbacks, we further explore the redundancy in the channel dimension and apply an architecture-level design with minor training costs. Therefore, we introduce CSKV, a training-efficient Channel Shrinking technique for KV cache compression: (1) We first analyze the singular value distribution of the KV cache, revealing significant redundancy and compression potential along the channel dimension. Based on this observation, we propose using low-rank decomposition for key and value layers and storing the low-dimension features. (2) To preserve model performance, we introduce a bi-branch KV cache, including a window-based full-precision KV cache and a low-precision compressed KV cache. (3) To reduce the training costs, we minimize the layer-wise reconstruction loss for the compressed KV cache instead of retraining the entire LLMs. Extensive experiments show that CSKV can reduce the memory overhead of the KV cache by 80% while maintaining the model's long-context capability. Moreover, we show that our method can be seamlessly combined with quantization to further reduce the memory overhead, achieving a compression ratio of up to 95%.
Authors: Haodi Wang, Tangyu Jiang, Yu Guo, Chengjun Cai, Cong Wang, Xiaohua Jia
Abstract: Deep learning models have been extensively adopted in various regions due to their ability to represent hierarchical features, which highly rely on the training set and procedures. Thus, protecting the training process and deep learning algorithms is paramount in privacy preservation. Although Differential Privacy (DP) as a powerful cryptographic primitive has achieved satisfying results in deep learning training, the existing schemes still fall short in preserving model utility, i.e., they either invoke a high noise scale or inevitably harm the original gradients. To address the above issues, in this paper, we present a more robust approach for DP training called GReDP. Specifically, we compute the model gradients in the frequency domain and adopt a new approach to reduce the noise level. Unlike the previous work, our GReDP only requires half of the noise scale compared to DPSGD [1] while keeping all the gradient information intact. We present a detailed analysis of our method both theoretically and empirically. The experimental results show that our GReDP works consistently better than the baselines on all models and training settings.
Authors: EverestAI, :, Sijing Chen, Yuan Feng, Laipeng He, Tianwei He, Wendi He, Yanni Hu, Bin Lin, Yiting Lin, Yu Pan, Pengfei Tan, Chengwei Tian, Chen Wang, Zhicheng Wang, Ruoye Xie, Jixun Yao, Quanlei Yan, Yuguang Yang, Jianhao Ye, Jingjing Yin, Yanzhen Yu, Huimin Zhang, Xiang Zhang, Guangcheng Zhao, Hongbin Zhou, Pengpeng Zou
Abstract: With the advent of the big data and large language model era, zero-shot personalized rapid customization has emerged as a significant trend. In this report, we introduce Takin AudioLLM, a series of techniques and models, mainly including Takin TTS, Takin VC, and Takin Morphing, specifically designed for audiobook production. These models are capable of zero-shot speech production, generating high-quality speech that is nearly indistinguishable from real human speech and facilitating individuals to customize the speech content according to their own needs. Specifically, we first introduce Takin TTS, a neural codec language model that builds upon an enhanced neural speech codec and a multi-task training framework, capable of generating high-fidelity natural speech in a zero-shot way. For Takin VC, we advocate an effective content and timbre joint modeling approach to improve the speaker similarity, while advocating for a conditional flow matching based decoder to further enhance its naturalness and expressiveness. Last, we propose the Takin Morphing system with highly decoupled and advanced timbre and prosody modeling approaches, which enables individuals to customize speech production with their preferred timbre and prosody in a precise and controllable manner. Extensive experiments validate the effectiveness and robustness of our Takin AudioLLM series models. For detailed demos, please refer to https://everest-ai.github.io/takinaudiollm/.
Authors: DaDong Jiang, Xianghui Yang, Zibo Zhao, Sheng Zhang, Jiaao Yu, Zeqiang Lai, Shaoxiong Yang, Chunchao Guo, Xiaobo Zhou, Zhihui Ke
Abstract: Recent texture generation methods achieve impressive results due to the powerful generative prior they leverage from large-scale text-to-image diffusion models. However, abstract textual prompts are limited in providing global textural or shape information, which results in the texture generation methods producing blurry or inconsistent patterns. To tackle this, we present FlexiTex, embedding rich information via visual guidance to generate a high-quality texture. The core of FlexiTex is the Visual Guidance Enhancement module, which incorporates more specific information from visual guidance to reduce ambiguity in the text prompt and preserve high-frequency details. To further enhance the visual guidance, we introduce a Direction-Aware Adaptation module that automatically designs direction prompts based on different camera poses, avoiding the Janus problem and maintaining semantically global consistency. Benefiting from the visual guidance, FlexiTex produces quantitatively and qualitatively sound results, demonstrating its potential to advance texture generation for real-world applications.
Authors: Junlin Lv, Yuan Feng, Xike Xie, Xin Jia, Qirong Peng, Guiming Xie
Abstract: Large language models have achieved notable success across various domains, yet efficient inference is still limited by the quadratic computation complexity of the attention mechanism. The inference consists of prefilling and decoding phases. Although several attempts have been made to accelerate decoding, the inefficiency of the prefilling phase, especially for long-context tasks, remains a challenge. In this paper, we observe a locality in query criticality during the prefilling phase of long-context processing: adjacent query tokens tend to focus on similar subsets of the past Key-Value (KV) cache. Based on this observation, we propose CritiPrefill, a criticality-based segment-wise prefilling method. This method partitions the input sequence's queries and KV cache into segments and blocks, utilizing a segment-wise algorithm to estimate the query criticality. By pruning non-critical computations between query segments and cache blocks in the self-attention mechanism, the prefilling process can be significantly accelerated. Extensive evaluations on multiple long-context datasets show up to 2.7x speedup on Llama3-8B and 3.0x speedup on Yi-9B for 128K context length on a single A100 GPU, with minimal quality degradation.
Authors: Thomas Savage, Stephen Ma, Abdessalem Boukil, Vishwesh Patel, Ekanath Rangan, Ivan Rodriguez, Jonathan H Chen
Abstract: Large Language Model (LLM) fine tuning is underutilized in the field of medicine. Two of the most common methods of fine tuning are Supervised Fine Tuning (SFT) and Direct Preference Optimization (DPO), but there is little guidance informing users when to use either technique. In this investigation, we compare the performance of SFT and DPO for five common natural language tasks in medicine: Classification with text data, Classification with numeric data, Clinical Reasoning, Summarization, and Clinical Triage. We find that SFT alone is sufficient for Classification with text data, whereas DPO improves performance for the more complex tasks of Clinical Reasoning, Summarization and Clinical Triage. Our results establish the role and importance of DPO fine tuning within medicine, and consequently call attention to current software gaps that prevent widespread deployment of this technique.
Authors: Hanyue Zhang, Zhiliu Yang, Xinhe Zuo, Yuxin Tong, Ying Long, Chen Liu
Abstract: This paper proposes a novel framework for large-scale scene reconstruction based on 3D Gaussian splatting (3DGS) and aims to address the scalability and accuracy challenges faced by existing methods. For tackling the scalability issue, we split the large scene into multiple cells, and the candidate point-cloud and camera views of each cell are correlated through a visibility-based camera selection and a progressive point-cloud extension. To reinforce the rendering quality, three highlighted improvements are made in comparison with vanilla 3DGS, which are a strategy of the ray-Gaussian intersection and the novel Gaussians density control for learning efficiency, an appearance decoupling module based on ConvKAN network to solve uneven lighting conditions in large-scale scenes, and a refined final loss with the color loss, the depth distortion loss, and the normal consistency loss. Finally, the seamless stitching procedure is executed to merge the individual Gaussian radiance field for novel view synthesis across different cells. Evaluation of Mill19, Urban3D, and MatrixCity datasets shows that our method consistently generates more high-fidelity rendering results than state-of-the-art methods of large-scale scene reconstruction. We further validate the generalizability of the proposed approach by rendering on self-collected video clips recorded by a commercial drone.
Authors: Shiyu Fang, Jiaqi Liu, Mingyu Ding, Yiming Cui, Chen Lv, Peng Hang, Jian Sun
Abstract: At present, Connected Autonomous Vehicles (CAVs) have begun to open road testing around the world, but their safety and efficiency performance in complex scenarios is still not satisfactory. Cooperative driving leverages the connectivity ability of CAVs to achieve synergies greater than the sum of their parts, making it a promising approach to improving CAV performance in complex scenarios. However, the lack of interaction and continuous learning ability limits current cooperative driving to single-scenario applications and specific Cooperative Driving Automation (CDA). To address these challenges, this paper proposes CoDrivingLLM, an interactive and learnable LLM-driven cooperative driving framework, to achieve all-scenario and all-CDA. First, since Large Language Models(LLMs) are not adept at handling mathematical calculations, an environment module is introduced to update vehicle positions based on semantic decisions, thus avoiding potential errors from direct LLM control of vehicle positions. Second, based on the four levels of CDA defined by the SAE J3216 standard, we propose a Chain-of-Thought (COT) based reasoning module that includes state perception, intent sharing, negotiation, and decision-making, enhancing the stability of LLMs in multi-step reasoning tasks. Centralized conflict resolution is then managed through a conflict coordinator in the reasoning process. Finally, by introducing a memory module and employing retrieval-augmented generation, CAVs are endowed with the ability to learn from their past experiences. We validate the proposed CoDrivingLLM through ablation experiments on the negotiation module, reasoning with different shots experience, and comparison with other cooperative driving methods.
Authors: Zishen Wan, Che-Kai Liu, Hanchen Yang, Ritik Raj, Chaojian Li, Haoran You, Yonggan Fu, Cheng Wan, Sixu Li, Youbin Kim, Ananda Samajdar, Yingyan Celine Lin, Mohamed Ibrahim, Jan M. Rabaey, Tushar Krishna, Arijit Raychowdhury
Abstract: The remarkable advancements in artificial intelligence (AI), primarily driven by deep neural networks, are facing challenges surrounding unsustainable computational trajectories, limited robustness, and a lack of explainability. To develop next-generation cognitive AI systems, neuro-symbolic AI emerges as a promising paradigm, fusing neural and symbolic approaches to enhance interpretability, robustness, and trustworthiness, while facilitating learning from much less data. Recent neuro-symbolic systems have demonstrated great potential in collaborative human-AI scenarios with reasoning and cognitive capabilities. In this paper, we aim to understand the workload characteristics and potential architectures for neuro-symbolic AI. We first systematically categorize neuro-symbolic AI algorithms, and then experimentally evaluate and analyze them in terms of runtime, memory, computational operators, sparsity, and system characteristics on CPUs, GPUs, and edge SoCs. Our studies reveal that neuro-symbolic models suffer from inefficiencies on off-the-shelf hardware, due to the memory-bound nature of vector-symbolic and logical operations, complex flow control, data dependencies, sparsity variations, and limited scalability. Based on profiling insights, we suggest cross-layer optimization solutions and present a hardware acceleration case study for vector-symbolic architecture to improve the performance, efficiency, and scalability of neuro-symbolic computing. Finally, we discuss the challenges and potential future directions of neuro-symbolic AI from both system and architectural perspectives.
Authors: Zhangchen Ye, Tao Jiang, Chenfeng Xu, Yiming Li, Hang Zhao
Abstract: Vision-based 3D occupancy prediction is significantly challenged by the inherent limitations of monocular vision in depth estimation. This paper introduces CVT-Occ, a novel approach that leverages temporal fusion through the geometric correspondence of voxels over time to improve the accuracy of 3D occupancy predictions. By sampling points along the line of sight of each voxel and integrating the features of these points from historical frames, we construct a cost volume feature map that refines current volume features for improved prediction outcomes. Our method takes advantage of parallax cues from historical observations and employs a data-driven approach to learn the cost volume. We validate the effectiveness of CVT-Occ through rigorous experiments on the Occ3D-Waymo dataset, where it outperforms state-of-the-art methods in 3D occupancy prediction with minimal additional computational cost. The code is released at \url{https://github.com/Tsinghua-MARS-Lab/CVT-Occ}.