Authors: Rose Bohrer, Ashe Neth
Abstract: Particularly in transgender and nonbinary (TGNB) communities, it is an increasingly common practice to publicly share one's personal pronouns so that we may be gendered correctly in others' speech. Many of us have nuanced desires for how we are gendered, leading us to use more complex descriptions of our wishes; for example, the descriptor 'she/they'. We observe that these descriptions of our wishes have the structure of a little language all their own. We thus propose formal logic as a tool for expressing one's personal pronouns and potentially other aspects of gender. We explore three potential logical foundations (linear logic, temporal logic, and free logic with definite descriptions) and their trade-offs. Our foremost motivation for this proposal is play, affirming that one can be both a logician and TGNB at the same time. We present formalization as something that can continue to evolve over time with society's understanding of gender. This implies that outreach is a major potential application: we can show TGNB youth that they belong in logic and have a unique contribution to make. Tools for evaluating whether one's pronouns are respected are an application as well.
Authors: Hongcheng Guo, Wei Zhang, Junhao Chen, Yaonan Gu, Jian Yang, Junjia Du, Binyuan Hui, Tianyu Liu, Jianxin Ma, Chang Zhou, Zhoujun Li
Abstract: Recently advancements in large multimodal models have led to significant strides in image comprehension capabilities. Despite these advancements, there is a lack of the robust benchmark specifically for assessing the Image-to-Web conversion proficiency of these large models. Primarily, it is essential to ensure the integrity of the web elements generated. These elements comprise visible and invisible categories. Previous evaluation methods (e.g., BLEU) are notably susceptible to significant alterations due to the presence of invisible elements in Web. Furthermore, it is crucial to measure the layout information of web pages, referring to the positional relationships between elements, which is overlooked by previous work. To address challenges, we have curated and aligned a benchmark of images and corresponding web codes (IW-Bench). Specifically, we propose the Element Accuracy, which tests the completeness of the elements by parsing the Document Object Model (DOM) tree. Layout Accuracy is also proposed to analyze the positional relationships of elements by converting DOM tree into a common subsequence. Besides, we design a five-hop multimodal Chain-of-Thought Prompting for better performance, which contains five hop: 1) SoM prompt injection. 2) Inferring Elements. 3) Inferring Layout. 4) Inferring Web code. 5) Reflection. Our benchmark comprises 1200 pairs of images and web codes with varying levels of difficulty. We have conducted extensive experiments on existing large multimodal models, offering insights into their performance and areas for improvement in image-to-web domain.
Authors: Yasser Saeid, Felix Neub\"urger, Stefanie Kr\"ugl, Helena H\"uster, Thomas Kopinski, Ralf Lanwehr
Abstract: This work investigates the identification of Charismatic Leadership Tactics (CLTs) in natural language using a fine-tuned Bidirectional Encoder Representations from Transformers (BERT) model. Based on an own extensive corpus of CLTs generated and curated for this task, our methodology entails training a machine learning model that is capable of accurately identifying the presence of these tactics in natural language. A performance evaluation is conducted to assess the effectiveness of our model in detecting CLTs. We find that the total accuracy over the detection of all CLTs is 98.96\% The results of this study have significant implications for research in psychology and management, offering potential methods to simplify the currently elaborate assessment of charisma in texts.
Authors: Xiaoyu Wang, Haoyong Ouyang, Balu Bhasuran, Xiao Luo, Karim Hanna, Mia Liza A. Lustria, Zhe He
Abstract: Accurate interpretation of lab results is crucial in clinical medicine, yet most patient portals use universal normal ranges, ignoring factors like age and gender. This study introduces Lab-AI, an interactive system that offers personalized normal ranges using Retrieval-Augmented Generation (RAG) from credible health sources. Lab-AI has two modules: factor retrieval and normal range retrieval. We tested these on 68 lab tests-30 with conditional factors and 38 without. For tests with factors, normal ranges depend on patient-specific information. Our results show that GPT-4-turbo with RAG achieved a 0.95 F1 score for factor retrieval and 0.993 accuracy for normal range retrieval. GPT-4-turbo with RAG outperformed the best non-RAG system by 29.1% in factor retrieval and showed 60.9% and 52.9% improvements in question-level and lab-level performance, respectively, for normal range retrieval. These findings highlight Lab-AI's potential to enhance patient understanding of lab results.
Authors: Xin Wang, Ting Dang, Vassilis Kostakos, Hong Jia
Abstract: Healthcare monitoring is crucial for early detection, timely intervention, and the ongoing management of health conditions, ultimately improving individuals' quality of life. Recent research shows that Large Language Models (LLMs) have demonstrated impressive performance in supporting healthcare tasks. However, existing LLM-based healthcare solutions typically rely on cloud-based systems, which raise privacy concerns and increase the risk of personal information leakage. As a result, there is growing interest in running these models locally on devices like mobile phones and wearables to protect users' privacy. Small Language Models (SLMs) are potential candidates to solve privacy and computational issues, as they are more efficient and better suited for local deployment. However, the performance of SLMs in healthcare domains has not yet been investigated. This paper examines the capability of SLMs to accurately analyze health data, such as steps, calories, sleep minutes, and other vital statistics, to assess an individual's health status. Our results show that, TinyLlama, which has 1.1 billion parameters, utilizes 4.31 GB memory, and has 0.48s latency, showing the best performance compared other four state-of-the-art (SOTA) SLMs on various healthcare applications. Our results indicate that SLMs could potentially be deployed on wearable or mobile devices for real-time health monitoring, providing a practical solution for efficient and privacy-preserving healthcare.
Authors: Xiang Li, Lan Zhao, Junhao Ren, Yajuan Sun, Chuan Fu Tan, Zhiquan Yeo, Gaoxi Xiao
Abstract: Effective information gathering and knowledge codification are pivotal for developing recommendation systems that promote circular economy practices. One promising approach involves the creation of a centralized knowledge repository cataloguing historical waste-to-resource transactions, which subsequently enables the generation of recommendations based on past successes. However, a significant barrier to constructing such a knowledge repository lies in the absence of a universally standardized framework for representing business activities across disparate geographical regions. To address this challenge, this paper leverages Large Language Models (LLMs) to classify textual data describing economic activities into the International Standard Industrial Classification (ISIC), a globally recognized economic activity classification framework. This approach enables any economic activity descriptions provided by businesses worldwide to be categorized into the unified ISIC standard, facilitating the creation of a centralized knowledge repository. Our approach achieves a 95% accuracy rate on a 182-label test dataset with fine-tuned GPT-2 model. This research contributes to the global endeavour of fostering sustainable circular economy practices by providing a standardized foundation for knowledge codification and recommendation systems deployable across regions.
Authors: Muhammad Junaid Khan, Gita Sukthankar
Abstract: This paper introduces SC-Phi2, a fine-tuned StarCraft II small language model for macromanagement tasks. Small language models, like Phi2, Gemma, and DistilBERT, are streamlined versions of large language models (LLMs) with fewer parameters that require less power and memory to run. To teach Microsoft's Phi2 model about StarCraft, we create a new SC2 text dataset with information about StarCraft races, roles, and actions and use it to fine-tune Phi-2 with self-supervised learning. We pair this language model with a Vision Transformer (ViT) from the pre-trained BLIP-2 (Bootstrapping Language Image Pre-training) model, fine-tuning it on the MSC replay dataset. This enables us to construct dynamic prompts that include visual game state information. Unlike the large models used in StarCraft LLMs such as GPT-3.5, Phi2 is trained primarily on textbook data and contains little inherent knowledge of StarCraft II beyond what is provided by our training process. By using LoRA (Low-rank Adaptation) and quantization, our model can be trained on a single GPU. We demonstrate that our model performs well at micromanagement tasks such as build order and global state prediction with a small number of parameters.
Authors: Ming Li, Keyu Chen, Ziqian Bi, Ming Liu, Benji Peng, Qian Niu, Junyu Liu, Jinlang Wang, Sen Zhang, Xuanhe Pan, Jiawei Xu, Pohsun Feng
Abstract: The rise of Multimodal Large Language Models (MLLMs) has become a transformative force in the field of artificial intelligence, enabling machines to process and generate content across multiple modalities, such as text, images, audio, and video. These models represent a significant advancement over traditional unimodal systems, opening new frontiers in diverse applications ranging from autonomous agents to medical diagnostics. By integrating multiple modalities, MLLMs achieve a more holistic understanding of information, closely mimicking human perception. As the capabilities of MLLMs expand, the need for comprehensive and accurate performance evaluation has become increasingly critical. This survey aims to provide a systematic review of benchmark tests and evaluation methods for MLLMs, covering key topics such as foundational concepts, applications, evaluation methodologies, ethical concerns, security, efficiency, and domain-specific applications. Through the classification and analysis of existing literature, we summarize the main contributions and methodologies of various surveys, conduct a detailed comparative analysis, and examine their impact within the academic community. Additionally, we identify emerging trends and underexplored areas in MLLM research, proposing potential directions for future studies. This survey is intended to offer researchers and practitioners a comprehensive understanding of the current state of MLLM evaluation, thereby facilitating further progress in this rapidly evolving field.
Authors: Tan T. Nguyen
Abstract: This review examines theoretical assumptions and computational models of event comprehension, tracing the evolution from discourse comprehension theories to contemporary event cognition frameworks. The review covers key discourse comprehension accounts, including Construction-Integration, Event Indexing, Causal Network, and Resonance models, highlighting their contributions to understanding cognitive processes in comprehension. I then discuss contemporary theoretical frameworks of event comprehension, including Event Segmentation Theory (Zacks et al., 2007), the Event Horizon Model (Radvansky & Zacks, 2014), and Hierarchical Generative Framework (Kuperberg, 2021), which emphasize prediction, causality, and multilevel representations in event understanding. Building on these theories, I evaluate five computational models of event comprehension: REPRISE (Butz et al., 2019), Structured Event Memory (SEM; Franklin et al., 2020), the Lu model (Lu et al., 2022), the Gumbsch model (Gumbsch et al., 2022), and the Elman and McRae model (2019). The analysis focuses on their approaches to hierarchical processing, prediction mechanisms, and representation learning. Key themes that emerge include the use of hierarchical structures as inductive biases, the importance of prediction in comprehension, and diverse strategies for learning event dynamics. The review identifies critical areas for future research, including the need for more sophisticated approaches to learning structured representations, integrating episodic memory mechanisms, and developing adaptive updating algorithms for working event models. By synthesizing insights from both theoretical frameworks and computational implementations, this review aims to advance our understanding of human event comprehension and guide future modeling efforts in cognitive science.
Authors: Isaac Kohane
Abstract: As large language models (LLMs) are deployed in high-stakes domains like healthcare, understanding how well their decision-making aligns with human preferences and values becomes crucial, especially when we recognize that there is no single gold standard for these preferences. This paper applies a systematic methodology for evaluating preference alignment in LLMs on categorical decision-making with medical triage as a domain-specific use case. It also measures how effectively an alignment procedure will change the alignment of a specific model. Key to this methodology is a novel simple measure, the Alignment Compliance Index (ACI), that quantifies how effectively a LLM can be aligned to a given preference function or gold standard. Since the ACI measures the effect rather than the process of alignment, it is applicable to alignment methods beyond the in-context learning used in this study. Using a dataset of simulated patient pairs, three frontier LLMs (GPT4o, Claude 3.5 Sonnet, and Gemini Advanced) were assessed on their ability to make triage decisions consistent with an expert clinician's preferences. The models' performance before and after alignment attempts was evaluated using various prompting strategies. The results reveal significant variability in alignment effectiveness across models and alignment approaches. Notably, models that performed well, as measured by ACI, pre-alignment sometimes degraded post-alignment, and small changes in the target preference function led to large shifts in model rankings. The implicit ethical principles, as understood by humans, underlying the LLMs' decisions were also explored through targeted questioning. This study motivates the use of a practical set of methods and the ACI, in the near term, to understand the correspondence between the variety of human and LLM decision-making values in categorical decision-making such as triage.
Authors: Shengsheng Qian, Zuyi Zhou, Dizhan Xue, Bing Wang, Changsheng Xu
Abstract: Cross-modal reasoning (CMR), the intricate process of synthesizing and drawing inferences across divergent sensory modalities, is increasingly recognized as a crucial capability in the progression toward more sophisticated and anthropomorphic artificial intelligence systems. Large Language Models (LLMs) represent a class of AI algorithms specifically engineered to parse, produce, and engage with human language on an extensive scale. The recent trend of deploying LLMs to tackle CMR tasks has marked a new mainstream of approaches for enhancing their effectiveness. This survey offers a nuanced exposition of current methodologies applied in CMR using LLMs, classifying these into a detailed three-tiered taxonomy. Moreover, the survey delves into the principal design strategies and operational techniques of prototypical models within this domain. Additionally, it articulates the prevailing challenges associated with the integration of LLMs in CMR and identifies prospective research directions. To sum up, this survey endeavors to expedite progress within this burgeoning field by endowing scholars with a holistic and detailed vista, showcasing the vanguard of current research whilst pinpointing potential avenues for advancement. An associated GitHub repository that collects the relevant papers can be found at https://github.com/ZuyiZhou/Awesome-Cross-modal-Reasoning-with-LLMs
URLs: https://github.com/ZuyiZhou/Awesome-Cross-modal-Reasoning-with-LLMs
Authors: Jiateng Liu, Lin Ai, Zizhou Liu, Payam Karisani, Zheng Hui, May Fung, Preslav Nakov, Julia Hirschberg, Heng Ji
Abstract: Propaganda plays a critical role in shaping public opinion and fueling disinformation. While existing research primarily focuses on identifying propaganda techniques, it lacks the ability to capture the broader motives and the impacts of such content. To address these challenges, we introduce propainsight, a conceptual framework grounded in foundational social science research, which systematically dissects propaganda into techniques, arousal appeals, and underlying intent. propainsight offers a more granular understanding of how propaganda operates across different contexts. Additionally, we present propagaze, a novel dataset that combines human-annotated data with high-quality synthetic data generated through a meticulously designed pipeline. Our experiments show that off-the-shelf LLMs struggle with propaganda analysis, but training with propagaze significantly improves performance. Fine-tuned Llama-7B-Chat achieves 203.4% higher text span IoU in technique identification and 66.2% higher BertScore in appeal analysis compared to 1-shot GPT-4-Turbo. Moreover, propagaze complements limited human-annotated data in data-sparse and cross-domain scenarios, showing its potential for comprehensive and generalizable propaganda analysis.
Authors: Mael Jullien, Alex Bogatu, Harriet Unsworth, Andre Freitas
Abstract: Matching patients to clinical trials demands a systematic and reasoned interpretation of documents which require significant expert-level background knowledge, over a complex set of well-defined eligibility criteria. Moreover, this interpretation process needs to operate at scale, over vast knowledge bases of trials. In this paper, we propose a scalable method that extends the capabilities of LLMs in the direction of systematizing the reasoning over sets of medical eligibility criteria, evaluating it in the context of real-world cases. The proposed method overlays a Set-guided reasoning method for LLMs. The proposed framework is evaluated on TREC 2022 Clinical Trials, achieving results superior to the state-of-the-art: NDCG@10 of 0.693 and Precision@10 of 0.73.
Authors: Graison Jos Thomas
Abstract: In the rapidly evolving field of financial sentiment analysis, the efficiency and accuracy of predictive models are critical due to their significant impact on financial markets. Transformer based models like BERT and large language models (LLMs) like GPT-4, have advanced NLP tasks considerably. Despite their advantages, BERT-based models face challenges with computational intensity in edge computing environments, and the substantial size and compute requirements of LLMs limit their practical deployment. This study proposes leveraging the generative capabilities of LLMs, such as GPT-4 Omni, to create synthetic, domain-specific training data. This approach addresses the challenge of data scarcity and enhances the performance of smaller models by making them competitive with their larger counterparts. The research specifically aims to enhance FinBERT, a BERT model fine-tuned for financial sentiment analysis, and develop TinyFinBERT, a compact transformer model, through a structured, two-tiered knowledge distillation strategy. Using data augmented by GPT-4 Omni, which involves generating new training examples and transforming existing data, we significantly improved the accuracy of FinBERT, preparing it to serve as a teacher model. This enhanced FinBERT then distilled knowledge to TinyFinBERT, employing both GPT-4 Omni and GPT-3.5 Turbo augmented data. The distillation strategy incorporated both logit and intermediate layer distillation. The training and evaluation of TinyFinBERT utilized the PhraseBank dataset and the FiQA 2018 Task1 dataset, achieving performance comparable to FinBERT while being substantially smaller and more efficient. This research demonstrates how LLMs can effectively contribute to the advancement of financial sentiment analysis by enhancing the capabilities of smaller, more efficient models through innovative data augmentation and distillation techniques.
Authors: Pedro Luiz Silva, Antonio de Domenico, Ali Maatouk, Fadhel Ayed
Abstract: Despite the remarkable success of Large Language Models (LLMs), they still exhibit a limited capability to align their outputs to the user instructions. In this work, we introduce a simple and effective method, which we name GUIDE, that mechanistically increases attention scores in instruction tokens. To support this operation, we present Influence, a novel metric that highlights how the user's instructions propagate through the transformer layers and impact the LLM output. Our results show that GUIDE improves the accuracy of following instructions 29.4 % to 60.4%, outperforming natural prompting alternatives and Supervised Fine-Tuning up to 1M tokens.
Authors: Mahmoud Abdelrahman, Edgardo Macatulad, Binyu Lei, Matias Quintana, Clayton Miller, Filip Biljecki
Abstract: The concept of digital twins has attracted significant attention across various domains, particularly within the built environment. However, there is a sheer volume of definitions and the terminological consensus remains out of reach. The lack of a universally accepted definition leads to ambiguities in their conceptualization and implementation, and may cause miscommunication for both researchers and practitioners. We employed Natural Language Processing (NLP) techniques to systematically extract and analyze definitions of digital twins from a corpus of 15,000 full-text articles spanning diverse disciplines in the built environment. The study compares these findings with insights from an expert survey that included 52 experts. The study identifies concurrence on the components that comprise a 'Digital Twin' from a practical perspective across various domains, contrasting them with those that do not, to identify deviations. We investigate the evolution of digital twin definitions over time and across different scales, including manufacturing, building, and urban/geospatial perspectives. We extracted the main components of Digital Twins using Text Frequency Analysis and N-gram analysis. Subsequently, we identified components that appeared in the literature and conducted a Chi-square test to assess the significance of each component in different domains. Our findings indicate that definitions differ based on the field of research in which they are conceived, but with many similarities across domains. One significant generalizable differentiation is related to whether a digital twin was used for High-Performance Real-Time (HPRT) or Long-Term Decision Support (LTDS) applications. We synthesized and contrasted the most representative definitions in each domain, culminating in a novel, data-driven definition specifically tailored for each context.
Authors: Sakhinana Sagar Srinivas, Vijay Sri Vaikunth, Venkataramana Runkana
Abstract: Patents are the currency of innovation, and like any currency, they need to be managed and protected (Gavin Potenza). Patents, as legal documents that secure intellectual property rights, play a critical role in technological innovation. The growing complexity of patent documents and the surge in patent applications have created a need for automated solutions in patent analysis. In this work, we present PatExpert, an autonomous multi-agent conversational framework designed to streamline and optimize patent-related tasks. The framework consists of a metaagent that coordinates task-specific expert agents for various patent-related tasks and a critique agent for error handling and feedback provision. The meta-agent orchestrates specialized expert agents, each fine-tuned for specific tasks such as patent classification, acceptance, claim generation, abstractive summarization, multi-patent analysis, and scientific hypothesis generation. For multi-patent analysis, the framework incorporates advanced methods like Graph Retrieval-Augmented Generation (GRAG) to enhance response accuracy and relevance by combining semantic similarity with knowledge graphs. Error handling is managed by critique agents (Gold-LLM-as-a-Judge and Reward-LLM-as-a-Judge), which evaluate output responses for accuracy and provide iterative feedback. The framework also prioritizes explainability, ensuring transparent justifications for decisions made during patent analysis. Its comprehensive capabilities make it a valuable tool for automating complex patent workflows, enhancing efficiency, accuracy, and compliance in patent-related tasks. Empirical evidence demonstrates significant improvements in patent processing tasks, concluding that the framework offers a robust solution for automating and optimizing patent analysis.
Authors: Liujianfu Wang, Yuyang Du, Jingqi Lin, Kexin Chen, Soung Chang Liew
Abstract: Large language models (LLMs) are being widely researched across various disciplines, with significant recent efforts focusing on adapting LLMs for understanding of how communication networks operate. However, over-reliance on prompting techniques hinders the full exploitation of the generalization ability of these models, and the lack of efficient fine-tuning methods prevents the full realization of lightweight LLMs' potential. This paper addresses these challenges by introducing our Rephrase and Contrast (RaC) framework, an efficient fine-tuning framework. RaC enhances LLMs' comprehension and critical thinking abilities by incorporating question reformulation and contrastive analysis of correct and incorrect answers during the fine-tuning process. Experimental results demonstrate a 63.73% accuracy improvement over the foundational model when tested on a comprehensive networking problem set. Moreover, to efficiently construct the dataset for RaC fine-tuning, we develop a GPT-assisted data mining method for generating high-quality question-answer (QA) pairs; furthermore, we introduce ChoiceBoost, a data augmentation technique that expands dataset size while reducing answer-order bias. Apart from these technical innovations, we contribute to the networking community by open-sourcing valuable research resources, including: 1) the fine-tuned networking model referred to as RaC-Net, 2) the training dataset used for fine-tuning the model, 3) three testing problem sets of different difficulties to serve as benchmarks for future research, and 4) code associated with the above resources.
Authors: Naman Goyal
Abstract: Recent improvement in large language models, open doors for certain new experiences for on-device applications which were not possible before. In this work, we propose 3 such new experiences in 2 categories. First we discuss experiences which can be powered in screen understanding i.e. understanding whats on user screen namely - (1) visual question answering, and (2) automated form filling based on previous screen. The second category of experience which can be extended are smart replies to support for multilingual speakers with code-switching. Code-switching occurs when a speaker alternates between two or more languages. To the best of our knowledge, this is first such work to propose these tasks and solutions to each of them, to bridge the gap between latest research and real world impact of the research in on-device applications.
Authors: Saumya Malik
Abstract: In this thesis, I evaluate the performance of Large Language Models (LLMs) on the Law School Admissions Test (LSAT), specifically the Logic Games section of the test. I focus on this section because it presents a complex logical reasoning task and thus is a valuable source of data for evaluating how modern, increasingly capable LLMs can handle hard logical reasoning tasks. I construct a dataset of LSAT logic games and their associated metadata, and extensively evaluate LLMs' performance in a Chain-of-Thought prompting setting. Given the weak performance in this setting, I explore other prompting frameworks on a smaller subset of the dataset, adapting ideas from Reflexion to this task. This results in a substantially improved accuracy of 70 percent for GPT-4 and 46 percent for GPT-3.5 on this data subset, highlighting the capacity of LLMs to revise their logical errors, despite initially weak performance. Finally, I analyze the types of logic games that models perform better or worse on, as well as the types of logical errors I observe from human annotation, providing detailed insights on the logical reasoning capabilities of LLMs.
Authors: Zhengpei Cheng, Yingyi Wu, Danyang Zhang, Jiacheng Hu, Yujian Long
Abstract: This study addresses the critical challenges of assessing foundational academic skills by leveraging advancements in natural language processing (NLP). Traditional assessment methods often struggle to provide timely and comprehensive feedback on key cognitive and linguistic aspects, such as coherence, syntax, and analytical reasoning. Our approach integrates multiple state-of-the-art NLP models, including BERT, RoBERTa, BART, DeBERTa, and T5, within an ensemble learning framework. These models are combined through stacking techniques using LightGBM and Ridge regression to enhance predictive accuracy. The methodology involves detailed data preprocessing, feature extraction, and pseudo-label learning to optimize model performance. By incorporating sophisticated NLP techniques and ensemble learning, this study significantly improves the accuracy and efficiency of assessments, offering a robust solution that surpasses traditional methods and opens new avenues for educational technology research focused on enhancing core academic competencies.
Authors: Heegyu Kim, Taeyang Jeon, Seunghwan Choi, Hyunsouk Cho
Abstract: Text-to-SQL technology has become crucial for translating natural language into SQL queries in various industries, enabling non-technical users to perform complex data operations. The need for accurate evaluation methods has increased as these systems have grown more sophisticated. However, we found that the Execution Accuracy (EX), the most promising evaluation metric, still shows a substantial portion of false positives and negatives compared to human evaluation. Thus, this paper introduces FLEX (False-Less EXecution), a novel approach to evaluating text-to-SQL systems using large language models (LLMs) to emulate human expert-level evaluation of SQL queries. Our method shows significantly higher agreement with human expert judgments, improving Cohen's kappa from 61 to 78.17. Re-evaluating top-performing models on the Spider and BIRD benchmarks using FLEX reveals substantial shifts in performance rankings, with an average performance decrease of 3.15 due to false positive corrections and an increase of 6.07 from addressing false negatives. This work contributes to a more accurate and nuanced evaluation of text-to-SQL systems, potentially reshaping our understanding of state-of-the-art performance in this field.
Authors: Krithiga Ramadass, Abrit Pal Singh, Srihari J, Sheetal Kalyani
Abstract: This work addresses the persistent challenges of substantial training time and GPU resource requirements even when training lightweight encoder-vocoder models for Textless NLP. We reduce training steps significantly while improving performance by a) leveraging learning rate schedulers for efficient and faster convergence b) optimizing hop length and c) tuning the interpolation scale factors for better audio quality. Additionally, we explore the latent space representation for Indian languages such as Tamil and Bengali for the acoustic unit discovery and voice conversion task. Our approach leverages a quantized encoder architecture, in conjunction with a vocoder which utilizes the proposed mixture of optimized hop length, tuned interpolation scale factors and a cyclic learning rate scheduler. We obtain consistently good results across English, Tamil and Bengali datasets. The proposed method excels in capturing complex linguistic patterns, resulting in clear reconstructed audio during voice conversion with significantly reduced training time.
Authors: Shangeetha Sivasothy, Scott Barnett, Stefanus Kurniawan, Zafaryab Rasool, Rajesh Vasa
Abstract: Retrieval Augmented Generation (RAG) is increasingly being used when building Generative AI applications. Evaluating these applications and RAG pipelines is mostly done manually, via a trial and error process. Automating evaluation of RAG pipelines requires overcoming challenges such as context misunderstanding, wrong format, incorrect specificity, and missing content. Prior works therefore focused on improving evaluation metrics as well as enhancing components within the pipeline using available question and answer datasets. However, they have not focused on 1) providing a schema for capturing different types of question-answer pairs or 2) creating a set of templates for generating question-answer pairs that can support automation of RAG pipeline evaluation. In this paper, we present a technique for generating variations in question-answer pairs to trigger failures in RAG pipelines. We validate 5 open-source RAG pipelines using 3 datasets. Our approach revealed the highest failure rates when prompts combine multiple questions: 91% for questions when spanning multiple documents and 78% for questions from a single document; indicating a need for developers to prioritise handling these combined questions. 60% failure rate was observed in academic domain dataset and 53% and 62% failure rates were observed in open-domain datasets. Our automated approach outperforms the existing state-of-the-art methods, by increasing the failure rate by 51% on average per dataset. Our work presents an automated approach for continuously monitoring the health of RAG pipelines, which can be integrated into existing CI/CD pipelines, allowing for improved quality.
Authors: Sathya Krishnan Suresh, Wu Mengjun, Tushar Pranav, Eng Siong Chng
Abstract: The scarcity of domain specific dialogue datasets across various domains, from academic topics to everyday conversations, limits the development of dialogue systems for various applications. Existing research is often constrained either by dialogue datasets that are too general or by niche domain dialogue datasets whose scale does not match the required scale for training dialogue systems. To address this gap, we introduce DiaSynth - a synthetic dialogue generation framework capable of generating high quality, contextually rich dialogues across a wide range of domains. Our approach differs from existing frameworks by dynamically generating dialogues that incorporate simulated personas, subtopics, and diverse conversational characteristics, using a Large Language Model (LLM) with Chain of Thought (CoT) reasoning to create contextually rich, domain-specific dialogues that closely mimic natural human interactions. DiaSynth produces tailored dialogues that emulate realistic conversations. We perform our experiments by generating synthetic data using different LLMs and few-shot examples from DialogSum and SAMSum. The pretrained language models fine-tuned on the synthetic data outperform the base models by 16.47%, while the comparison between models fine-tuned on in-domain data and synthetic data shows that the synthetic data is able to capture 90.48% of the distribution of the in-domain data. The quality of the data generated also scales with the size of LLMs. These results validate DiaSynth's potential as a robust alternative to traditional data collection methods.
Authors: Antonis Papasavva, Shane Johnson, Ed Lowther, Samantha Lundrigan, Enrico Mariconti, Anna Markovska, Nilufer Tuptuk
Abstract: Fraud is a prevalent offence that extends beyond financial loss, causing psychological and physical harm to victims. The advancements in online communication technologies alowed for online fraud to thrive in this vast network, with fraudsters increasingly using these channels for deception. With the progression of technologies like AI, there is a growing concern that fraud will scale up, using sophisticated methods, like deep-fakes in phishing campaigns, all generated by language generation models like ChatGPT. However, the application of AI in detecting and analyzing online fraud remains understudied. We conduct a Systematic Literature Review on AI and NLP techniques for online fraud detection. The review adhered the PRISMA-ScR protocol, with eligibility criteria including relevance to online fraud, use of text data, and AI methodologies. We screened 2,457 academic records, 350 met our eligibility criteria, and included 223. We report the state-of-the-art NLP techniques for analysing various online fraud categories; the training data sources; the NLP algorithms and models built; and the performance metrics employed for model evaluation. We find that current research on online fraud is divided into various scam activitiesand identify 16 different frauds that researchers focus on. This SLR enhances the academic understanding of AI-based detection methods for online fraud and offers insights for policymakers, law enforcement, and businesses on safeguarding against such activities. We conclude that focusing on specific scams lacks generalization, as multiple models are required for different fraud types. The evolving nature of scams limits the effectiveness of models trained on outdated data. We also identify issues in data limitations, training bias reporting, and selective presentation of metrics in model performance reporting, which can lead to potential biases in model evaluation.
Authors: Yan Liu, Xiaoyuan Yi, Xiaokang Chen, Jing Yao, Jingwei Yi, Daoguang Zan, Zheng Liu, Xing Xie, Tsung-Yi Ho
Abstract: The demand for regulating potentially risky behaviors of large language models (LLMs) has ignited research on alignment methods. Since LLM alignment heavily relies on reward models for optimization or evaluation, neglecting the quality of reward models may cause unreliable results or even misalignment. Despite the vital role reward models play in alignment, previous works have consistently overlooked their performance and used off-the-shelf reward models arbitrarily without verification, rendering the reward model ``\emph{an elephant in the room}''. To this end, this work first investigates the quality of the widely-used preference dataset, HH-RLHF, and curates a clean version, CHH-RLHF. Based on CHH-RLHF, we benchmark the accuracy of a broad range of reward models used in previous alignment works, unveiling the unreliability of using them both for optimization and evaluation. Furthermore, we systematically study the impact of reward model quality on alignment performance in three reward utilization paradigms. Extensive experiments reveal that better reward models perform as better human preference proxies. This work aims to awaken people to notice this huge elephant in alignment research. We call attention to the following issues: (1) The reward model needs to be rigorously evaluated, whether for alignment optimization or evaluation. (2) Considering the role of reward models, research efforts should not only concentrate on alignment algorithm, but also on developing more reliable human proxy.
Authors: Enrica Troiano, Sofie Labat, Marco Antonio Stranisci, Viviana Patti, Rossana Damiano, Roman Klinger
Abstract: There is a mismatch between psychological and computational studies on emotions. Psychological research aims at explaining and documenting internal mechanisms of these phenomena, while computational work often simplifies them into labels. Many emotion fundamentals remain under-explored in natural language processing, particularly how emotions develop and how people cope with them. To help reduce this gap, we follow theories on coping, and treat emotions as strategies to cope with salient situations (i.e., how people deal with emotion-eliciting events). This approach allows us to investigate the link between emotions and behavior, which also emerges in language. We introduce the task of coping identification, together with a corpus to do so, constructed via role-playing. We find that coping strategies realize in text even though they are challenging to recognize, both for humans and automatic systems trained and prompted on the same task. We thus open up a promising research direction to enhance the capability of models to better capture emotion mechanisms from text.
Authors: Aniket Deroy, Subhankar Maity
Abstract: Code generation by Llama 3.1 models, such as Meta's Llama 3.1 405B, represents a significant advancement in the field of artificial intelligence, particularly in natural language processing and programming automation. This paper explores the capabilities and applications of Llama-driven code generation, highlighting its ability to translate natural language prompts into executable code across multiple programming languages. Key features include contextual awareness, multi-language support, and enhanced debugging and optimization functionalities. By examining these aspects, we illustrate how Llama can serve as a versatile tool for developers of all skill levels, improving productivity and efficiency in software development. The potential implications for education, industry, and the future of coding practices are also discussed, underscoring the transformative impact of AI in programming. Experimentation shows that while Llama 3.1 405B performs well with simple algorithmic and data structure based problems, it still struggles with problems on Quantum Computing, Bioinformatics, and Artificial Intelligence.
Authors: Giordano d'Aloisio, Sophie Fortz, Carol Hanna, Daniel Fortunato, Avner Bensoussan, E\~naut Mendiluze Usandizaga, Federica Sarro
Abstract: Background: Quantum computing is a rapidly growing new programming paradigm that brings significant changes to the design and implementation of algorithms. Understanding quantum algorithms requires knowledge of physics and mathematics, which can be challenging for software developers. Aims: In this work, we provide a first analysis of how LLMs can support developers' understanding of quantum code. Method: We empirically analyse and compare the quality of explanations provided by three widely adopted LLMs (Gpt3.5, Llama2, and Tinyllama) using two different human-written prompt styles for seven state-of-the-art quantum algorithms. We also analyse how consistent LLM explanations are over multiple rounds and how LLMs can improve existing descriptions of quantum algorithms. Results: Llama2 provides the highest quality explanations from scratch, while Gpt3.5 emerged as the LLM best suited to improve existing explanations. In addition, we show that adding a small amount of context to the prompt significantly improves the quality of explanations. Finally, we observe how explanations are qualitatively and syntactically consistent over multiple rounds. Conclusions: This work highlights promising results, and opens challenges for future research in the field of LLMs for quantum code explanation. Future work includes refining the methods through prompt optimisation and parsing of quantum code explanations, as well as carrying out a systematic assessment of the quality of explanations.
Authors: Nikunj Saunshi, Stefani Karp, Shankar Krishnan, Sobhan Miryoosefi, Sashank J. Reddi, Sanjiv Kumar
Abstract: Given the increasing scale of model sizes, novel training strategies like gradual stacking [Gong et al., 2019, Reddi et al., 2023] have garnered interest. Stacking enables efficient training by gradually growing the depth of a model in stages and using layers from a smaller model in an earlier stage to initialize the next stage. Although efficient for training, the model biases induced by such growing approaches are largely unexplored. In this work, we examine this fundamental aspect of gradual stacking, going beyond its efficiency benefits. We propose a variant of gradual stacking called MIDAS that can speed up language model training by up to 40%. Furthermore we discover an intriguing phenomenon: MIDAS is not only training-efficient but surprisingly also has an inductive bias towards improving downstream tasks, especially tasks that require reasoning abilities like reading comprehension and math problems, despite having similar or slightly worse perplexity compared to baseline training. To further analyze this inductive bias, we construct reasoning primitives -- simple synthetic tasks that are building blocks for reasoning -- and find that a model pretrained with stacking is significantly better than standard pretraining on these primitives, with and without fine-tuning. This provides stronger and more robust evidence for this inductive bias towards reasoning. These findings of training efficiency and inductive bias towards reasoning are verified at 1B, 2B and 8B parameter language models. Finally, we conjecture the underlying reason for this inductive bias by exploring the connection of stacking to looped models and provide strong supporting empirical analysis.
Authors: Yu Fu, Jie He, Yifan Yang, Qun Liu, Deyi Xiong
Abstract: Meta learning has been widely used to exploit rich-resource source tasks to improve the performance of low-resource target tasks. Unfortunately, most existing meta learning approaches treat different source tasks equally, ignoring the relatedness of source tasks to the target task in knowledge transfer. To mitigate this issue, we propose a reinforcement-based multi-source meta-transfer learning framework (Meta-RTL) for low-resource commonsense reasoning. In this framework, we present a reinforcement-based approach to dynamically estimating source task weights that measure the contribution of the corresponding tasks to the target task in the meta-transfer learning. The differences between the general loss of the meta model and task-specific losses of source-specific temporal meta models on sampled target data are fed into the policy network of the reinforcement learning module as rewards. The policy network is built upon LSTMs that capture long-term dependencies on source task weight estimation across meta learning iterations. We evaluate the proposed Meta-RTL using both BERT and ALBERT as the backbone of the meta model on three commonsense reasoning benchmark datasets. Experimental results demonstrate that Meta-RTL substantially outperforms strong baselines and previous task selection strategies and achieves larger improvements on extremely low-resource settings.
Authors: Bryan Li, Aleksey Panasyuk, Chris Callison-Burch
Abstract: We study how differences in persuasive language across Wikipedia articles, written in either English and Russian, can uncover each culture's distinct perspective on different subjects. We develop a large language model (LLM) powered system to identify instances of persuasive language in multilingual texts. Instead of directly prompting LLMs to detect persuasion, which is subjective and difficult, we propose to reframe the task to instead ask high-level questions (HLQs) which capture different persuasive aspects. Importantly, these HLQs are authored by LLMs themselves. LLMs over-generate a large set of HLQs, which are subsequently filtered to a small set aligned with human labels for the original task. We then apply our approach to a large-scale, bilingual dataset of Wikipedia articles (88K total), using a two-stage identify-then-extract prompting strategy to find instances of persuasion. We quantify the amount of persuasion per article, and explore the differences in persuasion through several experiments on the paired articles. Notably, we generate rankings of articles by persuasion in both languages. These rankings match our intuitions on the culturally-salient subjects; Russian Wikipedia highlights subjects on Ukraine, while English Wikipedia highlights the Middle East. Grouping subjects into larger topics, we find politically-related events contain more persuasion than others. We further demonstrate that HLQs obtain similar performance when posed in either English or Russian. Our methodology enables cross-lingual, cross-cultural understanding at scale, and we release our code, prompts, and data.
Authors: Yulu Gan, Tomer Galanti, Tomaso Poggio, Eran Malach
Abstract: Originally proposed for handling time series data, Auto-regressive Decision Trees (ARDTs) have not yet been explored for language modeling. This paper delves into both the theoretical and practical applications of ARDTs in this new context. We theoretically demonstrate that ARDTs can compute complex functions, such as simulating automata, Turing machines, and sparse circuits, by leveraging "chain-of-thought" computations. Our analysis provides bounds on the size, depth, and computational efficiency of ARDTs, highlighting their surprising computational power. Empirically, we train ARDTs on simple language generation tasks, showing that they can learn to generate coherent and grammatically correct text on par with a smaller Transformer model. Additionally, we show that ARDTs can be used on top of transformer representations to solve complex reasoning tasks. This research reveals the unique computational abilities of ARDTs, aiming to broaden the architectural diversity in language model development.
Authors: Seth Aycock, David Stap, Di Wu, Christof Monz, Khalil Sima'an
Abstract: Extremely low-resource (XLR) languages lack substantial corpora for training NLP models, motivating the use of all available resources such as dictionaries and grammar books. Machine Translation from One Book (Tanzer et al., 2024) suggests prompting long-context LLMs with one grammar book enables English-Kalamang translation, an unseen XLR language - a noteworthy case of linguistic knowledge helping an NLP task. We investigate whether the book's grammatical explanations or its parallel examples are most effective for learning XLR translation, finding almost all improvement stems from the parallel examples. Further, we find similar results for Nepali, a seen low-resource language, and achieve performance comparable to an LLM with a grammar book by simply fine-tuning an encoder-decoder translation model. We then investigate where grammar books help by testing two linguistic tasks, grammaticality judgment and gloss prediction, and we explore what kind of grammatical knowledge helps by introducing a typological feature prompt that achieves leading results on these more relevant tasks. We thus emphasise the importance of task-appropriate data for XLR languages: parallel examples for translation, and grammatical data for linguistic tasks. As we find no evidence that long-context LLMs can make effective use of grammatical explanations for XLR translation, we suggest data collection for multilingual XLR tasks such as translation is best focused on parallel data over linguistic description.
Authors: Stefan Hackmann
Abstract: Foundation language model deployments often include auxiliary guard-rail models to filter or classify text, detecting jailbreak attempts, biased or toxic output, or ensuring topic adherence. These additional models increase the complexity and cost of model inference, especially since many are also large language models. To address this issue, we explore training-free model merging techniques to consolidate these models into a single, multi-functional model. We propose Heterogeneous Multi-Class Model Merging (HM3) as a simple technique for merging multi-class classifiers with heterogeneous label spaces. Unlike parameter-efficient fine-tuning techniques like LoRA, which require extensive training and add complexity during inference, recent advancements allow models to be merged in a training-free manner. We report promising results for merging BERT-based guard models, some of which attain an average F1-score higher than the source models while reducing the inference time by up to 44%. We introduce self-merging to assess the impact of reduced task-vector density, finding that the more poorly performing hate speech classifier benefits from self-merging while higher-performing classifiers do not, which raises questions about using task vector reduction for model tuning.
Authors: Ishani Mondal, Zongxia Li, Yufang Hou, Anandhavelu Natarajan, Aparna Garimella, Jordan Boyd-Graber
Abstract: Automating the creation of scientific diagrams from academic papers can significantly streamline the development of tutorials, presentations, and posters, thereby saving time and accelerating the process. Current text-to-image models struggle with generating accurate and visually appealing diagrams from long-context inputs. We propose SciDoc2Diagram, a task that extracts relevant information from scientific papers and generates diagrams, along with a benchmarking dataset, SciDoc2DiagramBench. We develop a multi-step pipeline SciDoc2Diagrammer that generates diagrams based on user intentions using intermediate code generation. We observed that initial diagram drafts were often incomplete or unfaithful to the source, leading us to develop SciDoc2Diagrammer-Multi-Aspect-Feedback (MAF), a refinement strategy that significantly enhances factual correctness and visual appeal and outperforms existing models on both automatic and human judgement.
Authors: Tatsuya Zetsu, Yuki Arase, Tomoyuki Kajiwara
Abstract: We propose edit operation based lexically constrained decoding for sentence simplification. In sentence simplification, lexical paraphrasing is one of the primary procedures for rewriting complex sentences into simpler correspondences. While previous studies have confirmed the efficacy of lexically constrained decoding on this task, their constraints can be loose and may lead to sub-optimal generation. We address this problem by designing constraints that replicate the edit operations conducted in simplification and defining stricter satisfaction conditions. Our experiments indicate that the proposed method consistently outperforms the previous studies on three English simplification corpora commonly used in this task.
Authors: Christine de Kock
Abstract: Research on extremist online communities frequently utilizes linguistic analysis to explore group dynamics and behaviour. Existing studies often rely on outdated lexicons that do not capture the evolving nature of in-group language, nor the social structure of the community. This paper proposes a novel method for inducing in-group lexicons which incorporates its socio-temporal context. Using dynamic word and user embeddings trained on conversations from online anti-women communities, our approach outperforms prior methods for lexicon induction. We provide a new lexicon of manosphere terms, validated by human experts, which quantifies the relevance of each term to a specific sub-community. We present novel insights on in-group language which illustrate the utility of this approach.
Authors: Jiwei Tang, Jin Xu, Tingwei Lu, Hai Lin, Yiming Zhao, Hai-Tao Zheng
Abstract: Large Language Models (LLMs) demonstrate exceptional capabilities in various scenarios. However, they suffer from much redundant information and tend to be lost in the middle in long context scenarios, leading to inferior performance. To address these challenges, we present Perception Compressor, a training-free prompt compression method. It includes a dual-slope ratio allocator to dynamically assign compression ratios and open-book ratios, a perception retriever that leverages guiding questions and instruction to retrieve the most relevant demonstrations, and a semi-guided iterative compression that retains key information at the token level while removing tokens that distract the LLM. We conduct extensive experiments on long context benchmarks, i.e., NaturalQuestions, LongBench, and MuSiQue. Experiment results show that Perception Compressor outperforms existing methods by a large margin, achieving state-of-the-art performance.
Authors: Haocheng Lin
Abstract: Large language models (LLMs) have transformed natural language processing across diverse fields, yet their general-purpose design limits their effectiveness in specialized domains, such as simulating opinions on environmental policies. This paper presents an approach for fine-tuning LLMs using data from the UK Household Longitudinal Study, improving the accuracy of opinion generation by conditioning models on socio-demographic factors like age, income, education, and region. By emulating diverse synthetic profiles, fine-tuned models capture the subtle differences across demographic groups more effectively than pre-trained versions. Metrics such as Chi-Squared, Cosine Similarity, Jaccard Index, and KL-divergence, demonstrate a strong alignment between synthetic and real-world opinion data. This approach highlights the potential of fine-tuning LLMs to provide more informed, representative, and ethical insights into public sentiments on environmental issues. The findings underscore the importance of tailoring LLMs to specific societal contexts for more accurate and ethical policy simulations.
Authors: Haowei Zhang, Jianzhe Liu, Zhen Han, Shuo Chen, Bailan He, Volker Tresp, Zhiqiang Xu, Jindong Gu
Abstract: Question decomposition has emerged as an effective strategy for prompting Large Language Models (LLMs) to answer complex questions. However, while existing methods primarily focus on unimodal language models, the question decomposition capability of Multimodal Large Language Models (MLLMs) has yet to be explored. To this end, this paper explores visual question decomposition on MLLMs. Specifically, we introduce a systematic evaluation framework including a dataset and several evaluation criteria to assess the quality of the decomposed sub-questions, revealing that existing MLLMs struggle to produce high-quality sub-questions. To address this limitation, we propose a specific finetuning dataset, DecoVQA+, for enhancing the model's question decomposition capability. Aiming at enabling models to perform appropriate selective decomposition, we propose an efficient finetuning pipeline. The finetuning pipeline consists of our proposed dataset and a training objective for selective decomposition. Finetuned MLLMs demonstrate significant improvements in the quality of sub-questions and the policy of selective question decomposition. Additionally, the models also achieve higher accuracy with selective decomposition on VQA benchmark datasets.
Authors: Xuyuan Xiong, Simeng Han, Ziyue Zhou, Arman Cohan
Abstract: Large Language Models (LLMs) are commonly used to generate solutions for mathematical reasoning problems in the following formats: natural language, code, or a combination of both. In this paper, we explore fundamental questions related to solving mathematical reasoning problems using natural language and code with state-of-the-art LLMs, including GPT-4o-mini and LLama-3.1-8b-Turbo. Our findings show that LLMs are better at reasoning in natural language compared to code. Additionally, although natural language and code serve as complementary forms of reasoning, they can affect each other in a negative way in certain scenarios. These insights motivate our development of a new prompting method, MetaMath, which leverages an LLM to dynamically select the most appropriate reasoning form, resulting in improved performance over comparable baselines with GPT-4o-mini.
Authors: Seongmin Lee, Jaewook Shin, Youngjin Ahn, Seokin Seo, Ohjoon Kwon, Kee-Eung Kim
Abstract: Recent advances in large language models (LLMs) have significantly impacted the domain of multi-hop question answering (MHQA), where systems are required to aggregate information and infer answers from disparate pieces of text. However, the autoregressive nature of LLMs inherently poses a challenge as errors may accumulate if mistakes are made in the intermediate reasoning steps. This paper introduces Monte-Carlo tree search for Zero-shot multi-hop Question Answering (MZQA), a framework based on Monte-Carlo tree search (MCTS) to identify optimal reasoning paths in MHQA tasks, mitigating the error propagation from sequential reasoning processes. Unlike previous works, we propose a zero-shot prompting method, which relies solely on instructions without the support of hand-crafted few-shot examples that typically require domain expertise. We also introduce a behavioral cloning approach (MZQA-BC) trained on self-generated MCTS inference trajectories, achieving an over 10-fold increase in reasoning speed with bare compromise in performance. The efficacy of our method is validated on standard benchmarks such as HotpotQA, 2WikiMultihopQA, and MuSiQue, demonstrating that it outperforms existing frameworks.
Authors: Zheng Wang, Zhongyang Li, Zeren Jiang, Dandan Tu, Wei Shi
Abstract: In the age of mobile internet, user data, often referred to as memories, is continuously generated on personal devices. Effectively managing and utilizing this data to deliver services to users is a compelling research topic. In this paper, we introduce a novel task of crafting personalized agents powered by large language models (LLMs), which utilize a user's smartphone memories to enhance downstream applications with advanced LLM capabilities. To achieve this goal, we introduce EMG-RAG, a solution that combines Retrieval-Augmented Generation (RAG) techniques with an Editable Memory Graph (EMG). This approach is further optimized using Reinforcement Learning to address three distinct challenges: data collection, editability, and selectability. Extensive experiments on a real-world dataset validate the effectiveness of EMG-RAG, achieving an improvement of approximately 10% over the best existing approach. Additionally, the personalized agents have been transferred into a real smartphone AI assistant, which leads to enhanced usability.
Authors: Dongyue Li, Ziniu Zhang, Lu Wang, Hongyang R. Zhang
Abstract: We study the problem of fine-tuning a language model (LM) for a target task by optimally using the information from $n$ auxiliary tasks. This problem has broad applications in NLP, such as targeted instruction tuning and data selection in chain-of-thought fine-tuning. The key challenge of this problem is that not all auxiliary tasks are useful to improve the performance of the target task. Thus, choosing the right subset of auxiliary tasks is crucial. Conventional subset selection methods, such as forward & backward selection, are unsuitable for LM fine-tuning because they require repeated training on subsets of auxiliary tasks. This paper introduces a new algorithm to estimate model fine-tuning performances without repeated training. Our algorithm first performs multitask training using the data of all the tasks to obtain a meta initialization. Then, we approximate the model fine-tuning loss of a subset using functional values and gradients from the meta initialization. Empirically, we find that this gradient-based approximation holds with remarkable accuracy for twelve transformer-based LMs. Thus, we can now estimate fine-tuning performances on CPUs within a few seconds. We conduct extensive experiments to validate our approach, delivering a speedup of $30\times$ over conventional subset selection while incurring only $1\%$ error of the true fine-tuning performances. In downstream evaluations of instruction tuning and chain-of-thought fine-tuning, our approach improves over prior methods that utilize gradient or representation similarity for subset selection by up to $3.8\%$.
Authors: Pablo Romero, Lifeng Han, Goran Nenadic
Abstract: Medication Extraction and Mining play an important role in healthcare NLP research due to its practical applications in hospital settings, such as their mapping into standard clinical knowledge bases (SNOMED-CT, BNF, etc.). In this work, we investigate state-of-the-art LLMs in text mining tasks on medications and their related attributes such as dosage, route, strength, and adverse effects. In addition, we explore different ensemble learning methods (\textsc{Stack-Ensemble} and \textsc{Voting-Ensemble}) to augment the model performances from individual LLMs. Our ensemble learning result demonstrated better performances than individually fine-tuned base models BERT, RoBERTa, RoBERTa-L, BioBERT, BioClinicalBERT, BioMedRoBERTa, ClinicalBERT, and PubMedBERT across general and specific domains. Finally, we build up an entity linking function to map extracted medical terminologies into the SNOMED-CT codes and the British National Formulary (BNF) codes, which are further mapped to the Dictionary of Medicines and Devices (dm+d), and ICD. Our model's toolkit and desktop applications are publicly available at \url{https://github.com/HECTA-UoM/ensemble-NER}.
Authors: Sachin Kumar
Abstract: LLMs(Large Language Models) nowadays have widespread adoption as a tool for solving issues across various domain/tasks. These models since are susceptible to produce harmful or toxic results, inference-time adversarial attacks, therefore they do undergo safety alignment training and Red teaming for putting in safety guardrails. For using these models, usually fine-tuning is done for model alignment on the desired tasks, which can make model more aligned but also make it more susceptible to produce unsafe responses, if fine-tuned with harmful data.In this paper, we study how much of impact introduction of harmful data in fine-tuning can make, and if it can override the safety protection of those models. Conversely,it was also explored that if model is fine-tuned on safety data can make the model produce more safer responses. Further we explore if fine-tuning the model on harmful data makes it less helpful or less trustworthy because of increase in model uncertainty leading to knowledge drift. Our extensive experimental results shown that Safety protection in an open-source can be overridden, when fine-tuned with harmful data as observed by ASR increasing by 35% when compared to basemodel's ASR. Also, as observed, fine-tuning a model with harmful data made the harmful fine-tuned model highly uncertain with huge knowledge drift and less truthfulness in its responses. Furthermore, for the safe fine-tuned model, ASR decreases by 51.68% as compared to the basemodel, and Safe model also shown in minor drop in uncertainty and truthfulness as compared to basemodel. This paper's code is available at: https://github.com/techsachinkr/Overriding_Model_Safety_Protections
URLs: https://github.com/techsachinkr/Overriding_Model_Safety_Protections
Authors: Ziyu Wang, Hao Li, Di Huang, Amir M. Rahmani
Abstract: In digital healthcare, large language models (LLMs) have primarily been utilized to enhance question-answering capabilities and improve patient interactions. However, effective patient care necessitates LLM chains that can actively gather information by posing relevant questions. This paper presents HealthQ, a novel framework designed to evaluate the questioning capabilities of LLM healthcare chains. We implemented several LLM chains, including Retrieval-Augmented Generation (RAG), Chain of Thought (CoT), and reflective chains, and introduced an LLM judge to assess the relevance and informativeness of the generated questions. To validate HealthQ, we employed traditional Natural Language Processing (NLP) metrics such as Recall-Oriented Understudy for Gisting Evaluation (ROUGE) and Named Entity Recognition (NER)-based set comparison, and constructed two custom datasets from public medical note datasets, ChatDoctor and MTS-Dialog. Our contributions are threefold: we provide the first comprehensive study on the questioning capabilities of LLMs in healthcare conversations, develop a novel dataset generation pipeline, and propose a detailed evaluation methodology.
Authors: Vibhor Agarwal, Yiqiao Jin, Mohit Chandra, Munmun De Choudhury, Srijan Kumar, Nishanth Sastry
Abstract: The remarkable capabilities of large language models (LLMs) in language understanding and generation have not rendered them immune to hallucinations. LLMs can still generate plausible-sounding but factually incorrect or fabricated information. As LLM-empowered chatbots become popular, laypeople may frequently ask health-related queries and risk falling victim to these LLM hallucinations, resulting in various societal and healthcare implications. In this work, we conduct a pioneering study of hallucinations in LLM-generated responses to real-world healthcare queries from patients. We propose MedHalu, a carefully crafted first-of-its-kind medical hallucination dataset with a diverse range of health-related topics and the corresponding hallucinated responses from LLMs with labeled hallucination types and hallucinated text spans. We also introduce MedHaluDetect framework to evaluate capabilities of various LLMs in detecting hallucinations. We also employ three groups of evaluators -- medical experts, LLMs, and laypeople -- to study who are more vulnerable to these medical hallucinations. We find that LLMs are much worse than the experts. They also perform no better than laypeople and even worse in few cases in detecting hallucinations. To fill this gap, we propose expert-in-the-loop approach to improve hallucination detection through LLMs by infusing expert reasoning. We observe significant performance gains for all the LLMs with an average macro-F1 improvement of 6.3 percentage points for GPT-4.
Authors: Aniket Pramanick, Yufang Hou, Saif M. Mohammad, Iryna Gurevych
Abstract: Natural Language Processing (NLP) is a dynamic, interdisciplinary field that integrates intellectual traditions from computer science, linguistics, social science, and more. Despite its established presence, the definition of what constitutes NLP research remains debated. In this work, we quantitatively investigate what constitutes NLP by examining research papers. For this purpose, we propose a taxonomy and introduce NLPContributions, a dataset of nearly $2k$ research paper abstracts, expertly annotated to identify scientific contributions and classify their types according to this taxonomy. We also propose a novel task to automatically identify these elements, for which we train a strong baseline on our dataset. We present experimental results from this task and apply our model to $\sim$$29k$ NLP research papers to analyze their contributions, aiding in the understanding of the nature of NLP research. Our findings reveal a rising involvement of machine learning in NLP since the early nineties, alongside a declining focus on adding knowledge about language or people; again, in post-2020, there has been a resurgence of focus on language and people. We hope this work will spark discussions on our community norms and inspire efforts to consciously shape the future.
Authors: Xiang Dai, Sarvnaz Karimi, Biaoyan Fang
Abstract: Effective summarisation evaluation metrics enable researchers and practitioners to compare different summarisation systems efficiently. Estimating the effectiveness of an automatic evaluation metric, termed meta-evaluation, is a critically important research question. In this position paper, we review recent meta-evaluation practices for summarisation evaluation metrics and find that (1) evaluation metrics are primarily meta-evaluated on datasets consisting of examples from news summarisation datasets, and (2) there has been a noticeable shift in research focus towards evaluating the faithfulness of generated summaries. We argue that the time is ripe to build more diverse benchmarks that enable the development of more robust evaluation metrics and analyze the generalization ability of existing evaluation metrics. In addition, we call for research focusing on user-centric quality dimensions that consider the generated summary's communicative goal and the role of summarisation in the workflow.
Authors: Aniket Pramanick, Yufang Hou, Saif M. Mohammad, Iryna Gurevych
Abstract: Large Language Models (LLMs) have ushered in a transformative era in Natural Language Processing (NLP), reshaping research and extending NLP's influence to other fields of study. However, there is little to no work examining the degree to which LLMs influence other research fields. This work empirically and systematically examines the influence and use of LLMs in fields beyond NLP. We curate $106$ LLMs and analyze $\sim$$148k$ papers citing LLMs to quantify their influence and reveal trends in their usage patterns. Our analysis reveals not only the increasing prevalence of LLMs in non-CS fields but also the disparities in their usage, with some fields utilizing them more frequently than others since 2018, notably Linguistics and Engineering together accounting for $\sim$$45\%$ of LLM citations. Our findings further indicate that most of these fields predominantly employ task-agnostic LLMs, proficient in zero or few-shot learning without requiring further fine-tuning, to address their domain-specific problems. This study sheds light on the cross-disciplinary impact of NLP through LLMs, providing a better understanding of the opportunities and challenges.
Authors: Yexing Du, Ziyang Ma, Yifan Yang, Keqi Deng, Xie Chen, Bo Yang, Yang Xiang, Ming Liu, Bing Qin
Abstract: Speech Language Models (SLMs) have demonstrated impressive performance on speech translation tasks. However, existing research primarily focuses on direct instruction fine-tuning and often overlooks the inherent reasoning capabilities of SLMs. In this paper, we introduce a three-stage training framework designed to activate the chain-of-thought (CoT) capabilities of SLMs. We propose CoT-ST, a speech translation model that utilizes multimodal CoT to decompose speech translation into sequential steps of speech recognition and translation. We validated the effectiveness of our method on two datasets: the CoVoST-2 dataset and MuST-C dataset. The experimental results demonstrate that CoT-ST outperforms previous state-of-the-art methods, achieving higher BLEU scores (CoVoST-2 en-ja: 30.5->30.8, en-zh: 45.2->47.7, MuST-C en-zh: 19.6->21.2). This work is open sourced at https://github.com/X-LANCE/SLAM-LLM/tree/main/examples/st_covost2 .
URLs: https://github.com/X-LANCE/SLAM-LLM/tree/main/examples/st_covost2
Authors: Shaolin Zhu, Leiyu Pan, Bo Li, Deyi Xiong
Abstract: Recent advancements in large language models (LLMs) have shown promising results in multilingual translation even with limited bilingual supervision. The major challenges are catastrophic forgetting and parameter interference for finetuning LLMs when provided parallel training data. To address these challenges, we propose LANDeRMT, a \textbf{L}anguage-\textbf{A}ware \textbf{N}euron \textbf{De}tecting and \textbf{R}outing framework that selectively finetunes LLMs to \textbf{M}achine \textbf{T}ranslation with diverse translation training data. In LANDeRMT, we evaluate the awareness of neurons to MT tasks and categorize them into language-general and language-specific neurons. This categorization enables selective parameter updates during finetuning, mitigating parameter interference and catastrophic forgetting issues. For the detected neurons, we further propose a conditional awareness-based routing mechanism to dynamically adjust language-general and language-specific capacity within LLMs, guided by translation signals. Experimental results demonstrate that the proposed LANDeRMT is very effective in learning translation knowledge, significantly improving translation quality over various strong baselines for multiple language pairs.
Authors: Siyuan Chen, Cong Ming, Zhiling Zhang, Yanyi Chen, Kenny Q. Zhu, Mengyue Wu
Abstract: In the realm of mental health support chatbots, it is vital to show empathy and encourage self-exploration to provide tailored solutions. However, current approaches tend to provide general insights or solutions without fully understanding the help-seeker's situation. Therefore, we propose PsyMix, a chatbot that integrates the analyses of the seeker's state from the perspective of a psychotherapy approach (Chain-of-Psychotherapies, CoP) before generating the response, and learns to incorporate the strength of various psychotherapies by fine-tuning on a mixture of CoPs. Through comprehensive evaluation, we found that PsyMix can outperform the ChatGPT baseline, and demonstrate a comparable level of empathy in its responses to that of human counselors.
Authors: Shahed Masoudian, Markus Frohman, Navid Rekabsaz, Markus Schedl
Abstract: Language models frequently inherit societal biases from their training data. Numerous techniques have been proposed to mitigate these biases during both the pre-training and fine-tuning stages. However, fine-tuning a pre-trained debiased language model on a downstream task can reintroduce biases into the model. Additionally, existing debiasing methods for downstream tasks either (i) require labels of protected attributes (e.g., age, race, or political views) that are often not available or (ii) rely on indicators of bias, which restricts their applicability to gender debiasing since they rely on gender-specific words. To address this, we introduce a novel debiasing regularization technique based on the class-wise variance of embeddings. Crucially, our method does not require attribute labels and targets any attribute, thus addressing the shortcomings of existing debiasing methods. Our experiments on encoder language models and three datasets demonstrate that our method outperforms existing strong debiasing baselines that rely on target attribute labels while maintaining performance on the target task.
Authors: Prakash Dhakal, Daya Sagar Baral
Abstract: Automatic text summarization in Nepali language is an unexplored area in natural language processing (NLP). Although considerable research has been dedicated to extractive summarization, the area of abstractive summarization, especially for low-resource languages such as Nepali, remains largely unexplored. This study explores the use of multilingual transformer models, specifically mBART and mT5, for generating headlines for Nepali news articles through abstractive summarization. The research addresses key challenges associated with summarizing texts in Nepali by first creating a summarization dataset through web scraping from various Nepali news portals. These multilingual models were then fine-tuned using different strategies. The performance of the fine-tuned models were then assessed using ROUGE scores and human evaluation to ensure the generated summaries were coherent and conveyed the original meaning. During the human evaluation, the participants were asked to select the best summary among those generated by the models, based on criteria such as relevance, fluency, conciseness, informativeness, factual accuracy, and coverage. During the evaluation with ROUGE scores, the 4-bit quantized mBART with LoRA model was found to be effective in generating better Nepali news headlines in comparison to other models and also it was selected 34.05% of the time during the human evaluation, outperforming all other fine-tuned models created for Nepali News headline generation.
Authors: Ante Wang, Linfeng Song, Zijun Min, Ge Xu, Xiaoli Wang, Junfeng Yao, Jinsong Su
Abstract: Conversational query generation aims at producing search queries from dialogue histories, which are then used to retrieve relevant knowledge from a search engine to help knowledge-based dialogue systems. Trained to maximize the likelihood of gold queries, previous models suffer from the data hunger issue, and they tend to both drop important concepts from dialogue histories and generate irrelevant concepts at inference time. We attribute these issues to the over-association phenomenon where a large number of gold queries are indirectly related to the dialogue topics, because annotators may unconsciously perform reasoning with their background knowledge when generating these gold queries. We carefully analyze the negative effects of this phenomenon on pretrained Seq2seq query producers and then propose effective instance-level weighting strategies for training to mitigate these issues from multiple perspectives. Experiments on two benchmarks, Wizard-of-Internet and DuSinc, show that our strategies effectively alleviate the negative effects and lead to significant performance gains (2%-5% across automatic metrics and human evaluation). Further analysis shows that our model selects better concepts from dialogue histories and is 10 times more data efficient than the baseline. The code is available at https://github.com/DeepLearnXMU/QG-OverAsso.
Authors: Gibong Hong, Veronica Hindle, Nadine M. Veasley, Hannah D. Holscher, Halil Kilicoglu
Abstract: Motivation: The gut microbiota has recently emerged as a key factor that underpins certain connections between diet and human health. A tremendous amount of knowledge has been amassed from experimental studies on diet, human metabolism and microbiome. However, this evidence remains mostly buried in scientific publications, and biomedical literature mining in this domain remains scarce. We developed DiMB-RE, a comprehensive corpus annotated with 15 entity types (e.g., Nutrient, Microorganism) and 13 relation types (e.g., increases, improves) capturing diet-microbiome associations. We also trained and evaluated state-of-the-art natural language processing (NLP) models for named entity, trigger, and relation extraction as well as factuality detection using DiMB-RE. Results: DiMB-RE consists of 14,450 entities and 4,206 relationships from 165 articles. While NLP models performed reasonably well for named entity recognition (0.760 F$_{1}$), end-to-end relation extraction performance was modest (0.356 F$_{1}$), partly due to missed entities and triggers as well as cross-sentence relations. Conclusions: To our knowledge, DiMB-RE is largest and most diverse dataset focusing on diet-microbiome interactions. It can serve as a benchmark corpus for biomedical literature mining. Availability: DiMB-RE and the NLP models are available at https://github.com/ScienceNLP-Lab/DiMB-RE.
Authors: Jialin Liu, Jianhua Wu, Jie Liu, Yutai Duan
Abstract: Fine-tuning large language models (LLMs) with Low-Rank adaption (LoRA) is widely acknowledged as an effective approach for continual learning for new tasks. However, it often suffers from catastrophic forgetting when dealing with multiple tasks sequentially. To this end, we propose Attentional Mixture of LoRAs (AM-LoRA), a continual learning approach tailored for LLMs. Specifically, AM-LoRA learns a sequence of LoRAs for a series of tasks to continually learn knowledge from different tasks. The key of our approach is that we devise an attention mechanism as a knowledge mixture module to adaptively integrate information from each LoRA. With the attention mechanism, AM-LoRA can efficiently leverage the distinctive contributions of each LoRA, while mitigating the risk of mutually negative interactions among them that may lead to catastrophic forgetting. Moreover, we further introduce $L1$ norm in the learning process to make the attention vector more sparse. The sparse constraints can enable the model to lean towards selecting a few highly relevant LoRAs, rather than aggregating and weighting all LoRAs collectively, which can further reduce the impact stemming from mutual interference. Experimental results on continual learning benchmarks indicate the superiority of our proposed method.
Authors: Maor Reuben, Ortal Slobodin, Aviad Elyshar, Idan-Chaim Cohen, Orna Braun-Lewensohn, Odeya Cohen, Rami Puzis
Abstract: Human-like personality traits have recently been discovered in large language models, raising the hypothesis that their (known and as yet undiscovered) biases conform with human latent psychological constructs. While large conversational models may be tricked into answering psychometric questionnaires, the latent psychological constructs of thousands of simpler transformers, trained for other tasks, cannot be assessed because appropriate psychometric methods are currently lacking. Here, we show how standard psychological questionnaires can be reformulated into natural language inference prompts, and we provide a code library to support the psychometric assessment of arbitrary models. We demonstrate, using a sample of 88 publicly available models, the existence of human-like mental health-related constructs (including anxiety, depression, and Sense of Coherence) which conform with standard theories in human psychology and show similar correlations and mitigation strategies. The ability to interpret and rectify the performance of language models by using psychological tools can boost the development of more explainable, controllable, and trustworthy models.
Authors: Fengzhu Zeng, Wenqian Li, Wei Gao, Yan Pang
Abstract: Detecting multimodal misinformation, especially in the form of image-text pairs, is crucial. Obtaining large-scale, high-quality real-world fact-checking datasets for training detectors is costly, leading researchers to use synthetic datasets generated by AI technologies. However, the generalizability of detectors trained on synthetic data to real-world scenarios remains unclear due to the distribution gap. To address this, we propose learning from synthetic data for detecting real-world multimodal misinformation through two model-agnostic data selection methods that match synthetic and real-world data distributions. Experiments show that our method enhances the performance of a small MLLM (13B) on real-world fact-checking datasets, enabling it to even surpass GPT-4V~\cite{GPT-4V}.
Authors: Xiaopeng Li, Shangwen Wang, Shezheng Song, Bin Ji, Huijun Liu, Shasha Li, Jun Ma, Jie Yu
Abstract: Knowledge editing has emerged as an efficient approach for updating the knowledge of large language models (LLMs), attracting increasing attention in recent research. However, there is a notable lack of effective measures to prevent the malicious misuse of this technology, which could lead to harmful edits in LLMs. These malicious modifications have the potential to cause LLMs to generate toxic content, misleading users into inappropriate actions. To address this issue, we introduce a novel task, \textbf{K}nowledge \textbf{E}diting \textbf{T}ype \textbf{I}dentification (KETI), aimed at identifying malicious edits in LLMs. As part of this task, we present KETIBench, a benchmark that includes five types of malicious updates and one type of benign update. Furthermore, we develop four classical classification models and three BERT-based models as baseline identifiers for both open-source and closed-source LLMs. Our experimental results, spanning 42 trials involving two models and three knowledge editing methods, demonstrate that all seven baseline identifiers achieve decent identification performance, highlighting the feasibility of identifying malicious edits in LLMs. Additional analyses reveal that the performance of the identifiers is independent of the efficacy of the knowledge editing methods and exhibits cross-domain generalization, enabling the identification of edits from unknown sources. All data and code are available in https://github.com/xpq-tech/KETI. Warning: This paper contains examples of toxic text.
Authors: Xin Li, Weize Chen, Qizhi Chu, Haopeng Li, Zhaojun Sun, Ran Li, Chen Qian, Yiwei Wei, Zhiyuan Liu, Chuan Shi, Maosong Sun, Cheng Yang
Abstract: The need to analyze graphs is ubiquitous across various fields, from social networks to biological research and recommendation systems. Therefore, enabling the ability of large language models (LLMs) to process graphs is an important step toward more advanced general intelligence. However, current LLM benchmarks on graph analysis require models to directly reason over the prompts describing graph topology, and are thus limited to small graphs with only a few dozens of nodes. In contrast, human experts typically write programs based on popular libraries for task solving, and can thus handle graphs with different scales. To this end, a question naturally arises: can LLMs analyze graphs like professionals? In this paper, we introduce ProGraph, a manually crafted benchmark containing 3 categories of graph tasks. The benchmark expects solutions based on programming instead of directly reasoning over raw inputs. Our findings reveal that the performance of current LLMs is unsatisfactory, with the best model achieving only 36% accuracy. To bridge this gap, we propose LLM4Graph datasets, which include crawled documents and auto-generated codes based on 6 widely used graph libraries. By augmenting closed-source LLMs with document retrieval and fine-tuning open-source ones on the codes, we show 11-32% absolute improvements in their accuracies. Our results underscore that the capabilities of LLMs in handling structured data are still under-explored, and show the effectiveness of LLM4Graph in enhancing LLMs' proficiency of graph analysis. The benchmark, datasets and enhanced open-source models are available at https://github.com/BUPT-GAMMA/ProGraph.
Authors: Chong Zhang, Yi Tu, Yixi Zhao, Chenshu Yuan, Huan Chen, Yue Zhang, Mingxu Chai, Ya Guo, Huijia Zhu, Qi Zhang, Tao Gui
Abstract: Modeling and leveraging layout reading order in visually-rich documents (VrDs) is critical in document intelligence as it captures the rich structure semantics within documents. Previous works typically formulated layout reading order as a permutation of layout elements, i.e. a sequence containing all the layout elements. However, we argue that this formulation does not adequately convey the complete reading order information in the layout, which may potentially lead to performance decline in downstream VrD tasks. To address this issue, we propose to model the layout reading order as ordering relations over the set of layout elements, which have sufficient expressive capability for the complete reading order information. To enable empirical evaluation on methods towards the improved form of reading order prediction (ROP), we establish a comprehensive benchmark dataset including the reading order annotation as relations over layout elements, together with a relation-extraction-based method that outperforms previous methods. Moreover, to highlight the practical benefits of introducing the improved form of layout reading order, we propose a reading-order-relation-enhancing pipeline to improve model performance on any arbitrary VrD task by introducing additional reading order relation inputs. Comprehensive results demonstrate that the pipeline generally benefits downstream VrD tasks: (1) with utilizing the reading order relation information, the enhanced downstream models achieve SOTA results on both two task settings of the targeted dataset; (2) with utilizing the pseudo reading order information generated by the proposed ROP model, the performance of the enhanced models has improved across all three models and eight cross-domain VrD-IE/QA task settings without targeted optimization.
Authors: Yiwei Li, Jiayi Shi, Shaoxiong Feng, Peiwen Yuan, Xinglin Wang, Boyuan Pan, Heda Wang, Yao Hu, Kan Li
Abstract: Instruction data is crucial for improving the capability of Large Language Models (LLMs) to align with human-level performance. Recent research LIMA demonstrates that alignment is essentially a process where the model adapts instructions' interaction style or format to solve various tasks, leveraging pre-trained knowledge and skills. Therefore, for instructional data, the most important aspect is the task it represents, rather than the specific semantics and knowledge information. The latent representations of instructions play roles for some instruction-related tasks like data selection and demonstrations retrieval. However, they are always derived from text embeddings, encompass overall semantic information that influences the representation of task categories. In this work, we introduce a new concept, instruction embedding, and construct Instruction Embedding Benchmark (IEB) for its training and evaluation. Then, we propose a baseline Prompt-based Instruction Embedding (PIE) method to make the representations more attention on tasks. The evaluation of PIE, alongside other embedding methods on IEB with two designed tasks, demonstrates its superior performance in accurately identifying task categories. Moreover, the application of instruction embeddings in four downstream tasks showcases its effectiveness and suitability for instruction-related tasks.
Authors: Nuowei Liu, Xinhao Chen, Hongyi Wu, Changzhi Sun, Man Lan, Yuanbin Wu, Xiaopeng Bai, Shaoguang Mao, Yan Xia
Abstract: Existing rhetorical understanding and generation datasets or corpora primarily focus on single coarse-grained categories or fine-grained categories, neglecting the common interrelations between different rhetorical devices by treating them as independent sub-tasks. In this paper, we propose the Chinese Essay Rhetoric Dataset (CERD), consisting of 4 commonly used coarse-grained categories including metaphor, personification, hyperbole and parallelism and 23 fine-grained categories across both form and content levels. CERD is a manually annotated and comprehensive Chinese rhetoric dataset with five interrelated sub-tasks. Unlike previous work, our dataset aids in understanding various rhetorical devices, recognizing corresponding rhetorical components, and generating rhetorical sentences under given conditions, thereby improving the author's writing proficiency and language usage skills. Extensive experiments are conducted to demonstrate the interrelations between multiple tasks in CERD, as well as to establish a benchmark for future research on rhetoric. The experimental results indicate that Large Language Models achieve the best performance across most tasks, and jointly fine-tuning with multiple tasks further enhances performance.
Authors: Jia-Nan Li, Jian Guan, Wei Wu, Zhengtao Yu, Rui Yan
Abstract: Tables are ubiquitous across various domains for concisely representing structured information. Empowering large language models (LLMs) to reason over tabular data represents an actively explored direction. However, since typical LLMs only support one-dimensional~(1D) inputs, existing methods often flatten the two-dimensional~(2D) table structure into a sequence of tokens, which can severely disrupt the spatial relationships and result in an inevitable loss of vital contextual information. In this paper, we first empirically demonstrate the detrimental impact of such flattening operations on the performance of LLMs in capturing the spatial information of tables through two elaborate proxy tasks. Subsequently, we introduce a simple yet effective positional encoding method, termed ``2D-TPE'' (Two-Dimensional Table Positional Encoding), to address this challenge. 2D-TPE enables each attention head to dynamically select a permutation order of tokens within the context for attending to them, where each permutation represents a distinct traversal mode for the table, such as column-wise or row-wise traversal. 2D-TPE effectively mitigates the risk of losing essential spatial information while preserving computational efficiency, thus better preserving the table structure. Extensive experiments across five benchmarks demonstrate that 2D-TPE outperforms strong baselines, underscoring the importance of preserving the table structure for accurate table comprehension. Comprehensive analysis further reveals the substantially better scalability of 2D-TPE to large tables than baselines.
Authors: Hyungjoo Chae, Taeyoon Kwon, Seungjun Moon, Yongho Song, Dongjin Kang, Kai Tzu-iunn Ong, Beong-woo Kwak, Seonghyeon Bae, Seung-won Hwang, Jinyoung Yeo
Abstract: This paper presents Coffee-Gym, a comprehensive RL environment for training models that provide feedback on code editing. Coffee-Gym includes two major components: (1) Coffee, a dataset containing humans' code edit traces for coding questions and machine-written feedback for editing erroneous code; (2) CoffeeEval, a reward function that faithfully reflects the helpfulness of feedback by assessing the performance of the revised code in unit tests. With them, Coffee-Gym addresses the unavailability of high-quality datasets for training feedback models with RL, and provides more accurate rewards than the SOTA reward model (i.e., GPT-4). By applying Coffee-Gym, we elicit feedback models that outperform baselines in enhancing open-source code LLMs' code editing, making them comparable with closed-source LLMs. We make the dataset and the model checkpoint publicly available.
Authors: Lei Sun, Jinming Zhao, Qin Jin
Abstract: Personality recognition aims to identify the personality traits implied in user data such as dialogues and social media posts. Current research predominantly treats personality recognition as a classification task, failing to reveal the supporting evidence for the recognized personality. In this paper, we propose a novel task named Explainable Personality Recognition, aiming to reveal the reasoning process as supporting evidence of the personality trait. Inspired by personality theories, personality traits are made up of stable patterns of personality state, where the states are short-term characteristic patterns of thoughts, feelings, and behaviors in a concrete situation at a specific moment in time. We propose an explainable personality recognition framework called Chain-of-Personality-Evidence (CoPE), which involves a reasoning process from specific contexts to short-term personality states to long-term personality traits. Furthermore, based on the CoPE framework, we construct an explainable personality recognition dataset from dialogues, PersonalityEvd. We introduce two explainable personality state recognition and explainable personality trait recognition tasks, which require models to recognize the personality state and trait labels and their corresponding support evidence. Our extensive experiments based on Large Language Models on the two tasks show that revealing personality traits is very challenging and we present some insights for future research. Our data and code are available at https://github.com/Lei-Sun-RUC/PersonalityEvd.
Authors: Jonathan Bourne
Abstract: OCR errors are common in digitised historical archives significantly affecting their usability and value. Generative Language Models (LMs) have shown potential for correcting these errors using the context provided by the corrupted text and the broader socio-cultural context, a process called Context Leveraging OCR Correction (CLOCR-C). However, getting sufficient training data for fine-tuning such models can prove challenging. This paper shows that fine-tuning a language model on synthetic data using an LM and using a character level Markov corruption process can significantly improve the ability to correct OCR errors. Models trained on synthetic data reduce the character error rate by 55% and word error rate by 32% over the base LM and outperform models trained on real data. Key findings include; training on under-corrupted data is better than over-corrupted data; non-uniform character level corruption is better than uniform corruption; More tokens-per-observation outperforms more observations for a fixed token budget. The outputs for this paper are a set of 8 heuristics for training effective CLOCR-C models, a dataset of 11,000 synthetic 19th century newspaper articles and scrambledtext a python library for creating synthetic corrupted data.
Authors: Lotem Peled-Cohen, Roi Reichart
Abstract: The close link between cognitive decline and language has fostered long-standing collaboration between the NLP and medical communities in dementia research. To examine this, we reviewed over 200 papers applying NLP to dementia related efforts, drawing from medical, technological, and NLP-focused literature. We identify key research areas, including dementia detection, linguistic biomarker extraction, caregiver support, and patient assistance, showing that half of all papers focus solely on dementia detection using clinical data. However, many directions remain unexplored: artificially degraded language models, synthetic data, digital twins, and more. We highlight gaps and opportunities around trust, scientific rigor, applicability, and cross-community collaboration, and showcase the diverse datasets encountered throughout our review: recorded, written, structured, spontaneous, synthetic, clinical, social media based, and more. This review aims to inspire more creative approaches to dementia research within the medical and NLP communities.
Authors: Tao Tan, Yining Qian, Ang Lv, Hongzhan Lin, Songhao Wu, Yongbo Wang, Feng Wang, Jingtong Wu, Xin Lu, Rui Yan
Abstract: Large language models (LLMs) enhanced with retrieval-augmented generation (RAG) have introduced a new paradigm for web search. However, the limited context awareness of LLMs degrades their performance on RAG tasks. Existing methods to enhance context awareness are often inefficient, incurring time or memory overhead during inference, and many are tailored to specific position embeddings. In this paper, we propose Position-Embedding-Agnostic attention Re-weighting (PEAR), which enhances the context awareness of LLMs with zero inference overhead. Specifically, on a proxy task focused on context copying, we first detect heads which suppress the models' context awareness thereby diminishing RAG performance. To weaken the impact of these heads, we re-weight their outputs with learnable coefficients. The LLM (with frozen parameters) is optimized by adjusting these coefficients to minimize loss on the proxy task. As a result, the coefficients are optimized to values less than one, thereby reducing their tendency to suppress RAG performance. During inference, the optimized coefficients are fixed to re-weight these heads, regardless of the specific task at hand. Our proposed PEAR offers two major advantages over previous approaches: (1) It introduces zero additional inference overhead in terms of memory usage or inference time, while outperforming competitive baselines in accuracy and efficiency across various RAG tasks. (2) It is independent of position embedding algorithms, ensuring broader applicability.
Authors: Enamul Hoque, Mohammed Saidul Islam
Abstract: Natural language and visualization are two complementary modalities of human communication that play a crucial role in conveying information effectively. While visualizations help people discover trends, patterns, and anomalies in data, natural language descriptions help explain these insights. Thus, combining text with visualizations is a prevalent technique for effectively delivering the core message of the data. Given the rise of natural language generation (NLG), there is a growing interest in automatically creating natural language descriptions for visualizations, which can be used as chart captions, answering questions about charts, or telling data-driven stories. In this survey, we systematically review the state of the art on NLG for visualizations and introduce a taxonomy of the problem. The NLG tasks fall within the domain of Natural Language Interfaces (NLI) for visualization, an area that has garnered significant attention from both the research community and industry. To narrow down the scope of the survey, we primarily concentrate on the research works that focus on text generation for visualizations. To characterize the NLG problem and the design space of proposed solutions, we pose five Wh-questions, why and how NLG tasks are performed for visualizations, what the task inputs and outputs are, as well as where and when the generated texts are integrated with visualizations. We categorize the solutions used in the surveyed papers based on these "five Wh-questions." Finally, we discuss the key challenges and potential avenues for future research in this domain.
Authors: Duy-Tung Pham, Thien Trang Nguyen Vu, Tung Nguyen, Linh Ngo Van, Duc Anh Nguyen, Thien Huu Nguyen
Abstract: Recent advances in neural topic models have concentrated on two primary directions: the integration of the inference network (encoder) with a pre-trained language model (PLM) and the modeling of the relationship between words and topics in the generative model (decoder). However, the use of large PLMs significantly increases inference costs, making them less practical for situations requiring low inference times. Furthermore, it is crucial to simultaneously model the relationships between topics and words as well as the interrelationships among topics themselves. In this work, we propose a novel framework called NeuroMax (Neural Topic Model with Maximizing Mutual Information with Pretrained Language Model and Group Topic Regularization) to address these challenges. NeuroMax maximizes the mutual information between the topic representation obtained from the encoder in neural topic models and the representation derived from the PLM. Additionally, NeuroMax employs optimal transport to learn the relationships between topics by analyzing how information is transported among them. Experimental results indicate that NeuroMax reduces inference time, generates more coherent topics and topic groups, and produces more representative document embeddings, thereby enhancing performance on downstream tasks.
Authors: Yike Wu, Yi Huang, Nan Hu, Yuncheng Hua, Guilin Qi, Jiaoyan Chen, Jeff Z. Pan
Abstract: Recent studies have explored the use of Large Language Models (LLMs) with Retrieval Augmented Generation (RAG) for Knowledge Graph Question Answering (KGQA). They typically require rewriting retrieved subgraphs into natural language formats comprehensible to LLMs. However, when tackling complex questions, the knowledge rewritten by existing methods may include irrelevant information, omit crucial details, or fail to align with the question's semantics. To address them, we propose a novel rewriting method CoTKR, Chain-of-Thought Enhanced Knowledge Rewriting, for generating reasoning traces and corresponding knowledge in an interleaved manner, thereby mitigating the limitations of single-step knowledge rewriting. Additionally, to bridge the preference gap between the knowledge rewriter and the question answering (QA) model, we propose a training strategy PAQAF, Preference Alignment from Question Answering Feedback, for leveraging feedback from the QA model to further optimize the knowledge rewriter. We conduct experiments using various LLMs across several KGQA benchmarks. Experimental results demonstrate that, compared with previous knowledge rewriting methods, CoTKR generates the most beneficial knowledge representation for QA models, which significantly improves the performance of LLMs in KGQA.
Authors: Yung-Chieh Chan, George Pu, Apaar Shanker, Parth Suresh, Penn Jenks, John Heyer, Sam Denton
Abstract: As large language models (LLMs) are applied to more use cases, creating high quality, task-specific datasets for fine-tuning becomes a bottleneck for model improvement. Using high quality human data has been the most common approach to unlock model performance, but is prohibitively expensive in many scenarios. Several alternative methods have also emerged, such as generating synthetic or hybrid data, but the effectiveness of these approaches remain unclear, especially in resource-constrained scenarios and tasks that are not easily verified. To investigate this, we group various synthetic data generation strategies into three representative categories -- Answer Augmentation, Question Rephrase and New Question -- and study the performance of student LLMs trained under various constraints, namely seed instruction set size and query budget. We demonstrate that these strategies are not equally effective across settings. Notably, the optimal data generation strategy depends strongly on the ratio between the available teacher query budget and the size of the seed instruction set. When this ratio is low, generating new answers to existing questions proves most effective, but as this ratio increases, generating new questions becomes optimal. Across all tasks, we find that choice of augmentation method and other design choices matter substantially more in low to mid data regimes than in high data regimes. We provide a practical framework for selecting the appropriate augmentation method across settings, taking into account additional factors such as the scalability of each method, the importance of verifying synthetic data, and the use of different LLMs for synthetic data generation.
Authors: Son Quoc Tran, Matt Kretchmar
Abstract: This paper proposes a novel training method to improve the robustness of Extractive Question Answering (EQA) models. Previous research has shown that existing models, when trained on EQA datasets that include unanswerable questions, demonstrate a significant lack of robustness against distribution shifts and adversarial attacks. Despite this, the inclusion of unanswerable questions in EQA training datasets is essential for ensuring real-world reliability. Our proposed training method includes a novel loss function for the EQA problem and challenges an implicit assumption present in numerous EQA datasets. Models trained with our method maintain in-domain performance while achieving a notable improvement on out-of-domain datasets. This results in an overall F1 score improvement of 5.7 across all testing sets. Furthermore, our models exhibit significantly enhanced robustness against two types of adversarial attacks, with a performance decrease of only about a third compared to the default models.
Authors: Hyunwoo Yoo
Abstract: Pre-trained language models such as DNABERT2 and Nucleotide Transformer, which are trained on DNA sequences, have shown promising performance in DNA sequence classification tasks. The classification ability of these models stems from language models trained on vast amounts of DNA sequence samples, followed by fine-tuning with relatively smaller classification datasets. However, these text-based systems are not robust enough and can be vulnerable to adversarial examples. While adversarial attacks have been widely studied in text classification, there is limited research in DNA sequence classification. In this paper, we adapt commonly used attack algorithms in text classification for DNA sequence classification. We evaluated the impact of various attack methods on DNA sequence classification at the character, word, and sentence levels. Our findings indicate that actual DNA language model sequence classifiers are vulnerable to these attacks.
Authors: Hongyi Yuan, Sheng Yu
Abstract: Electronic medical records (EMRs) contain the majority of patients' healthcare details. It is an abundant resource for developing an automatic healthcare system. Most of the natural language processing (NLP) studies on EMR processing, such as concept extraction, are adversely affected by the inaccurate segmentation of EMR sections. At the same time, not enough attention has been given to the accurate sectioning of EMRs. The information that may occur in section structures is unvalued. This work focuses on the segmentation of EMRs and proposes a black-box segmentation method using a simple sentence embedding model and neural network, along with a proper training method. To achieve universal adaptivity, we train our model on the dataset with different section headings formats. We compare several advanced deep learning-based NLP methods, and our method achieves the best segmentation accuracies (above 98%) on various test data with a proper training corpus.
Authors: Xuyang Wu, Shuowei Li, Hsin-Tai Wu, Zhiqiang Tao, Yi Fang
Abstract: RAG (Retrieval-Augmented Generation) have recently gained significant attention for their enhanced ability to integrate external knowledge sources in open-domain question answering (QA) tasks. However, it remains unclear how these models address fairness concerns, particularly with respect to sensitive attributes such as gender, geographic location, and other demographic factors. First, as language models evolve to prioritize utility, like improving exact match accuracy, fairness may have been largely overlooked. Second, RAG methods are complex pipelines, making it hard to identify and address biases, as each component is optimized for different goals. In this paper, we aim to empirically evaluate fairness in several RAG methods. We propose a fairness evaluation framework tailored to RAG methods, using scenario-based questions and analyzing disparities across demographic attributes. The experimental results indicate that, despite recent advances in utility-driven optimization, fairness issues persist in both the retrieval and generation stages, highlighting the need for more targeted fairness interventions within RAG pipelines. We will release our dataset and code upon acceptance of the paper.
Authors: Haoyu Zhao, Simran Kaur, Dingli Yu, Anirudh Goyal, Sanjeev Arora
Abstract: As large language models (LLMs) become increasingly advanced, their ability to exhibit compositional generalization -- the capacity to combine learned skills in novel ways not encountered during training -- has garnered significant attention. This type of generalization, particularly in scenarios beyond training data, is also of great interest in the study of AI safety and alignment. A recent study introduced the SKILL-MIX evaluation, where models are tasked with composing a short paragraph demonstrating the use of a specified $k$-tuple of language skills. While small models struggled with composing even with $k=3$, larger models like GPT-4 performed reasonably well with $k=5$ and $6$. In this paper, we employ a setup akin to SKILL-MIX to evaluate the capacity of smaller models to learn compositional generalization from examples. Utilizing a diverse set of language skills -- including rhetorical, literary, reasoning, theory of mind, and common sense -- GPT-4 was used to generate text samples that exhibit random subsets of $k$ skills. Subsequent fine-tuning of 7B and 13B parameter models on these combined skill texts, for increasing values of $k$, revealed the following findings: (1) Training on combinations of $k=2$ and $3$ skills results in noticeable improvements in the ability to compose texts with $k=4$ and $5$ skills, despite models never having seen such examples during training. (2) When skill categories are split into training and held-out groups, models significantly improve at composing texts with held-out skills during testing despite having only seen training skills during fine-tuning, illustrating the efficacy of the training approach even with previously unseen skills. This study also suggests that incorporating skill-rich (potentially synthetic) text into training can substantially enhance the compositional capabilities of models.
Authors: Tom\'a\v{s} Musil, David Mare\v{c}ek
Abstract: Large language models follow a lineage of many NLP applications that were directly inspired by distributional semantics, but do not seem to be closely related to it anymore. In this paper, we propose to employ the distributional theory of meaning once again. Using Independent Component Analysis to overcome some of its challenging aspects, we show that large language models represent semantic features in their hidden states.
Authors: Masanori Hirano, Kentaro Imajo
Abstract: This paper proposes a novel method for constructing instruction-tuned large language models (LLMs) for finance without instruction data. Traditionally, developing such domain-specific LLMs has been resource-intensive, requiring a large dataset and significant computational power for continual pretraining and instruction tuning. Our study proposes a simpler approach that combines domain-specific continual pretraining with model merging. Given that general-purpose pretrained LLMs and their instruction-tuned LLMs are often publicly available, they can be leveraged to obtain the necessary instruction task vector. By merging this with a domain-specific pretrained vector, we can effectively create instruction-tuned LLMs for finance without additional instruction data. Our process involves two steps: first, we perform continual pretraining on financial data; second, we merge the instruction-tuned vector with the domain-specific pretrained vector. Our experiments demonstrate the successful construction of instruction-tuned LLMs for finance. One major advantage of our method is that the instruction-tuned and domain-specific pretrained vectors are nearly independent. This independence makes our approach highly effective. The Japanese financial instruction-tuned LLMs we developed in this study are available at https://huggingface.co/pfnet/nekomata-14b-pfn-qfin-inst-merge.
URLs: https://huggingface.co/pfnet/nekomata-14b-pfn-qfin-inst-merge.
Authors: Huangyu Dai, Ben Chen, Kaidi Chen, Ying Han, Zihan Liang, Wen Jiang
Abstract: For crosslingual conversation and trade, Neural Machine Translation (NMT) is pivotal yet faces persistent challenges with monotony and repetition in generated content. Traditional solutions that rely on penalizing text redundancy or token reoccurrence have shown limited efficacy, particularly for lengthy article and e-commerce descriptions with inherent redundancy, even with the advent of Large Language Models (LLMs). This paper investigates the underlying causes of textual repetition through the lens of information entropy, attributing the phenomenon to the elevated uncertainty within the input text. To address this, a novel algorithm named Contrastive Token Learning with Similarity Decay (CTSD) is introduced, which modulates the suppression of tokens dynamically, informed by varying attention weights and inter-token distances. Furthermore, an e-commerce dataset comprised of title texts of online real items is compiled and released susceptible to hallucination translations to benchmark the algorithm. Extensive evaluations demonstrate that CTSD significantly outperforms existing approaches in precision and generalizability. Additional online A/B testing underscores its practical value, showing marked improvements in user engagement and conversion. Notably, this method has been implemented with full traffic on eight multilingual sites of alibaba.com, the largest B2B e-commerce platform in the world.
Authors: Yuho Lee, Taewon Lee, Jason Cai, Hang Su, Hwanjun Song
Abstract: Existing benchmarks for summarization quality evaluation often lack diverse input scenarios, focus on narrowly defined dimensions (e.g., faithfulness), and struggle with subjective and coarse-grained annotation schemes. To address these shortcomings, we create UniSumEval benchmark, which extends the range of input context (e.g., domain, length) and provides fine-grained, multi-dimensional annotations. We use AI assistance in data creation, identifying potentially hallucinogenic input texts, and also helping human annotators reduce the difficulty of fine-grained annotation tasks. With UniSumEval, we benchmark nine latest language models as summarizers, offering insights into their performance across varying input contexts and evaluation dimensions. Furthermore, we conduct a thorough comparison of SOTA automated summary evaluators. Our benchmark data will be available at https://github.com/DISL-Lab/UniSumEval-v1.0.
Authors: Ming Li, Ziqian Bi, Tianyang Wang, Keyu Chen, Jiawei Xu, Qian Niu, Junyu Liu, Benji Peng, Sen Zhang, Xuanhe Pan, Jinlang Wang, Pohsun Feng, Caitlyn Heqi Yin, Yizhu Wen, Ming Liu
Abstract: Object-Oriented Programming (OOP) has become a crucial paradigm for managing the growing complexity of modern software systems, particularly in fields like machine learning, deep learning, large language models (LLM), and data analytics. This work provides a comprehensive introduction to the integration of OOP techniques within these domains, with a focus on improving code modularity, maintainability, and scalability. We begin by outlining the evolution of computing and the rise of OOP, followed by an in-depth discussion of key OOP principles such as encapsulation, inheritance, polymorphism, and abstraction. The practical application of these principles is demonstrated using Python, a widely adopted language in AI and data science. Furthermore, we examine how design patterns and modular programming can be employed to enhance the structure and efficiency of machine learning systems. In subsequent sections, we apply these OOP concepts to real-world AI tasks, including the encapsulation of preprocessing workflows, machine learning model training, and evaluation. Detailed examples illustrate how OOP can be used to build reusable, scalable machine learning systems while maintaining code clarity and reducing redundancy.This work is intended to serve as a bridge for both beginners and experienced developers, equipping them with the necessary knowledge to apply OOP methodologies in AI-driven projects, ultimately fostering the development of more robust and maintainable systems.
Authors: Momose Oyama, Hiroaki Yamagiwa, Hidetoshi Shimodaira
Abstract: Independent Component Analysis (ICA) is an effective method for interpreting the intrinsic geometric structure of embeddings as semantic components. While ICA theory assumes that embeddings can be linearly decomposed into independent components, real-world data often do not satisfy this assumption. Consequently, there are remaining non-independencies between the estimated components that ICA cannot eliminate. We quantified these non-independencies using higher-order correlations and demonstrated that when the higher-order correlation between two components is large, it indicates a strong semantic association between them. The entire structure was revealed through visualization using a maximum spanning tree of semantic components. These findings allow for further understanding of embeddings through ICA.
Authors: Masato Fujitake
Abstract: In this paper, we create benchmarks and assess the effectiveness of error correction methods for Japanese vouchers in OCR (Optical Character Recognition) systems. It is essential for automation processing to correctly recognize scanned voucher text, such as the company name on invoices. However, perfect recognition is complex due to the noise, such as stamps. Therefore, it is crucial to correctly rectify erroneous OCR results. However, no publicly available OCR error correction benchmarks for Japanese exist, and methods have not been adequately researched. In this study, we measured text recognition accuracy by existing services on Japanese vouchers and developed a post-OCR correction benchmark. Then, we proposed simple baselines for error correction using language models and verified whether the proposed method could effectively correct these errors. In the experiments, the proposed error correction algorithm significantly improved overall recognition accuracy.
Authors: Eitan Wagner, Yuli Slavutsky, Omri Abend
Abstract: Although language model scores are often treated as probabilities, their reliability as probability estimators has mainly been studied through calibration, overlooking other aspects. In particular, it is unclear whether language models produce the same value for different ways of assigning joint probabilities to word spans. Our work introduces a novel framework, ConTestS (Consistency Testing over Spans), involving statistical tests to assess score consistency across interchangeable completion and conditioning orders. We conduct experiments on post-release real and synthetic data to eliminate training effects. Our findings reveal that both Masked Language Models (MLMs) and autoregressive models exhibit inconsistent predictions, with autoregressive models showing larger discrepancies. Larger MLMs tend to produce more consistent predictions, while autoregressive models show the opposite trend. Moreover, for both model types, prediction entropies offer insights into the true word span likelihood and therefore can aid in selecting optimal decoding strategies. The inconsistencies revealed by our analysis, as well their connection to prediction entropies and differences between model types, can serve as useful guides for future research on addressing these limitations.
Authors: Zhe Li, Wei Zhao, Yige Li, Jun Sun
Abstract: Influence functions aim to quantify the impact of individual training data points on a model's predictions. While extensive research has been conducted on influence functions in traditional machine learning models, their application to large language models (LLMs) has been limited. In this work, we conduct a systematic study to address a key question: do influence functions work on LLMs? Specifically, we evaluate influence functions across multiple tasks and find that they consistently perform poorly in most settings. Our further investigation reveals that their poor performance can be attributed to: (1) inevitable approximation errors when estimating the iHVP component due to the scale of LLMs, (2) uncertain convergence during fine-tuning, and, more fundamentally, (3) the definition itself, as changes in model parameters do not necessarily correlate with changes in LLM behavior. Our study thus suggests the need for alternative approaches for identifying influential samples. To support future work, our code is made available at https://github.com/plumprc/Failures-of-Influence-Functions-in-LLMs.
URLs: https://github.com/plumprc/Failures-of-Influence-Functions-in-LLMs.
Authors: Haoyu Zhang, Wenbin Wang, Tianshu Yu
Abstract: The field of Multimodal Sentiment Analysis (MSA) has recently witnessed an emerging direction seeking to tackle the issue of data incompleteness. Recognizing that the language modality typically contains dense sentiment information, we consider it as the dominant modality and present an innovative Language-dominated Noise-resistant Learning Network (LNLN) to achieve robust MSA. The proposed LNLN features a dominant modality correction (DMC) module and dominant modality based multimodal learning (DMML) module, which enhances the model's robustness across various noise scenarios by ensuring the quality of dominant modality representations. Aside from the methodical design, we perform comprehensive experiments under random data missing scenarios, utilizing diverse and meaningful settings on several popular datasets (\textit{e.g.,} MOSI, MOSEI, and SIMS), providing additional uniformity, transparency, and fairness compared to existing evaluations in the literature. Empirically, LNLN consistently outperforms existing baselines, demonstrating superior performance across these challenging and extensive evaluation metrics.
Authors: Menna Fateen, Bo Wang, Tsunenori Mine
Abstract: Automatic short answer scoring (ASAS) helps reduce the grading burden on educators but often lacks detailed, explainable feedback. Existing methods in ASAS with feedback (ASAS-F) rely on fine-tuning language models with limited datasets, which is resource-intensive and struggles to generalize across contexts. Recent approaches using large language models (LLMs) have focused on scoring without extensive fine-tuning. However, they often rely heavily on prompt engineering and either fail to generate elaborated feedback or do not adequately evaluate it. In this paper, we propose a modular retrieval augmented generation based ASAS-F system that scores answers and generates feedback in strict zero-shot and few-shot learning scenarios. We design our system to be adaptable to various educational tasks without extensive prompt engineering using an automatic prompt generation framework. Results show an improvement in scoring accuracy by 9\% on unseen questions compared to fine-tuning, offering a scalable and cost-effective solution.
Authors: Marios Kerasiotis, Loukas Ilias, Dimitris Askounis
Abstract: The detection of depression in social media posts is crucial due to the increasing prevalence of mental health issues. Traditional machine learning algorithms often fail to capture intricate textual patterns, limiting their effectiveness in identifying depression. Existing studies have explored various approaches to this problem but often fall short in terms of accuracy and robustness. To address these limitations, this research proposes a neural network architecture leveraging transformer-based models combined with metadata and linguistic markers. The study employs DistilBERT, extracting information from the last four layers of the transformer, applying learned weights, and averaging them to create a rich representation of the input text. This representation, augmented by metadata and linguistic markers, enhances the model's comprehension of each post. Dropout layers prevent overfitting, and a Multilayer Perceptron (MLP) is used for final classification. Data augmentation techniques, inspired by the Easy Data Augmentation (EDA) methods, are also employed to improve model performance. Using BERT, random insertion and substitution of phrases generate additional training data, focusing on balancing the dataset by augmenting underrepresented classes. The proposed model achieves weighted Precision, Recall, and F1-scores of 84.26%, 84.18%, and 84.15%, respectively. The augmentation techniques significantly enhance model performance, increasing the weighted F1-score from 72.59% to 84.15%.
Authors: Luka Andren\v{s}ek, Boshko Koloski, Andra\v{z} Pelicon, Nada Lavra\v{c}, Senja Pollak, Matthew Purver
Abstract: We investigate zero-shot cross-lingual news sentiment detection, aiming to develop robust sentiment classifiers that can be deployed across multiple languages without target-language training data. We introduce novel evaluation datasets in several less-resourced languages, and experiment with a range of approaches including the use of machine translation; in-context learning with large language models; and various intermediate training regimes including a novel task objective, POA, that leverages paragraph-level information. Our results demonstrate significant improvements over the state of the art, with in-context learning generally giving the best performance, but with the novel POA approach giving a competitive alternative with much lower computational overhead. We also show that language similarity is not in itself sufficient for predicting the success of cross-lingual transfer, but that similarity in semantic content and structure can be equally important.
Authors: Hippolyte Gisserot-Boukhlef, Ricardo Rei, Emmanuel Malherbe, C\'eline Hudelot, Pierre Colombo, Nuno M. Guerreiro
Abstract: Neural metrics for machine translation (MT) evaluation have become increasingly prominent due to their superior correlation with human judgments compared to traditional lexical metrics. Researchers have therefore utilized neural metrics through quality-informed decoding strategies, achieving better results than likelihood-based methods. With the rise of Large Language Models (LLMs), preference-based alignment techniques have gained attention for their potential to enhance translation quality by optimizing model weights directly on preferences induced by quality estimators. This study focuses on Contrastive Preference Optimization (CPO) and conducts extensive experiments to evaluate the impact of preference-based alignment on translation quality. Our findings indicate that while CPO consistently outperforms Supervised Fine-Tuning (SFT) on high-quality data with regard to the alignment metric, it may lead to instability across downstream evaluation metrics, particularly between neural and lexical ones. Additionally, we demonstrate that relying solely on the base model for generating candidate translations achieves performance comparable to using multiple external systems, while ensuring better consistency across downstream metrics.
Authors: Kaisi Guan, Qian Cao, Yuchong Sun, Xiting Wang, Ruihua Song
Abstract: Retrieval Augmented Generation (RAG) system is important in domains such as e-commerce, which has many long-tail entities and frequently updated information. Most existing works adopt separate modules for retrieval and generation, which may be suboptimal since the retrieval task and the generation task cannot benefit from each other to improve performance. We propose a novel Backbone Shared RAG framework (BSharedRAG). It first uses a domain-specific corpus to continually pre-train a base model as a domain-specific backbone model and then trains two plug-and-play Low-Rank Adaptation (LoRA) modules based on the shared backbone to minimize retrieval and generation losses respectively. Experimental results indicate that our proposed BSharedRAG outperforms baseline models by 5% and 13% in Hit@3 upon two datasets in retrieval evaluation and by 23% in terms of BLEU-3 in generation evaluation. Our codes, models, and dataset are available at https://bsharedrag.github.io.
Authors: Zining Zhang, Yao Chen, Bingsheng He, Zhenjie Zhang
Abstract: The increasing size and complexity of Large Language Models (LLMs) pose challenges for their deployment on personal computers and mobile devices. Aggressive post-training model compression is necessary to reduce the models' size, but it often results in significant accuracy loss. To address this challenge, we propose a novel network pruning technology that utilizes over 0.7 sparsity and less than 8 bits of quantization. Our approach enables the compression of prevailing LLMs within a couple of hours while maintaining a relatively small accuracy loss. In experimental evaluations, our method demonstrates effectiveness and potential for practical deployment. By making LLMs available on domestic devices, our work can facilitate a new era of natural language processing applications with wide-ranging impacts.
Authors: Jonathan D. Thomas, Andrea Silvi, Devdatt Dubhashi, Vikas Garg, Moa Johansson
Abstract: A central but unresolved aspect of problem-solving in AI is the capability to introduce and use abstractions, something humans excel at. Work in cognitive science has demonstrated that humans tend towards higher levels of abstraction when engaged in collaborative task-oriented communication, enabling gradually shorter and more information-efficient utterances. Several computational methods have attempted to replicate this phenomenon, but all make unrealistic simplifying assumptions about how abstractions are introduced and learned. Our method, Abstractions for Communicating Efficiently (ACE), overcomes these limitations through a neuro-symbolic approach. On the symbolic side, we draw on work from library learning for proposing abstractions. We combine this with neural methods for communication and reinforcement learning, via a novel use of bandit algorithms for controlling the exploration and exploitation trade-off in introducing new abstractions. ACE exhibits similar tendencies to humans on a collaborative construction task from the cognitive science literature, where one agent (the architect) instructs the other (the builder) to reconstruct a scene of block-buildings. ACE results in the emergence of an efficient language as a by-product of collaborative communication. Beyond providing mechanistic insights into human communication, our work serves as a first step to providing conversational agents with the ability for human-like communicative abstractions.
Authors: Vincent Beliveau, Helene Kaas, Martin Prener, Claes N. Ladefoged, Desmond Elliott, Gitte M. Knudsen, Lars H. Pinborg, Melanie Ganz
Abstract: Natural language processing (NLP) in the medical domain can underperform in real-world applications involving small datasets in a non-English language with few labeled samples and imbalanced classes. There is yet no consensus on how to approach this problem. We evaluated a set of NLP models including BERT-like transformers, few-shot learning with sentence transformers (SetFit), and prompted large language models (LLM), using three datasets of radiology reports on magnetic resonance images of epilepsy patients in Danish, a low-resource language. Our results indicate that BERT-like models pretrained in the target domain of radiology reports currently offer the optimal performances for this scenario. Notably, the SetFit and LLM models underperformed compared to BERT-like models, with LLM performing the worst. Importantly, none of the models investigated was sufficiently accurate to allow for text classification without any supervision. However, they show potential for data filtering, which could reduce the amount of manual labeling required.
Authors: Chanjun Park, Hyunsoo Ha, Jihoo Kim, Yungi Kim, Dahyun Kim, Sukyung Lee, Seonghoon Yang
Abstract: In this paper, we propose the 1 Trillion Token Platform (1TT Platform), a novel framework designed to facilitate efficient data sharing with a transparent and equitable profit-sharing mechanism. The platform fosters collaboration between data contributors, who provide otherwise non-disclosed datasets, and a data consumer, who utilizes these datasets to enhance their own services. Data contributors are compensated in monetary terms, receiving a share of the revenue generated by the services of the data consumer. The data consumer is committed to sharing a portion of the revenue with contributors, according to predefined profit-sharing arrangements. By incorporating a transparent profit-sharing paradigm to incentivize large-scale data sharing, the 1TT Platform creates a collaborative environment to drive the advancement of NLP and LLM technologies.
Authors: Aswathy Velutharambath, Amelie W\"uhrl, Roman Klinger
Abstract: The statement "The earth is flat" is factually inaccurate, but if someone truly believes and argues in its favor, it is not deceptive. Research on deception detection and fact checking often conflates factual accuracy with the truthfulness of statements. This assumption makes it difficult to (a) study subtle distinctions and interactions between the two and (b) gauge their effects on downstream tasks. The belief-based deception framework disentangles these properties by defining texts as deceptive when there is a mismatch between what people say and what they truly believe. In this study, we assess if presumed patterns of deception generalize to German language texts. We test the effectiveness of computational models in detecting deception using an established corpus of belief-based argumentation. Finally, we gauge the impact of deception on the downstream task of fact checking and explore if this property confounds verification models. Surprisingly, our analysis finds no correlation with established cues of deception. Previous work claimed that computational models can outperform humans in deception detection accuracy, however, our experiments show that both traditional and state-of-the-art models struggle with the task, performing no better than random guessing. For fact checking, we find that Natural Language Inference-based verification performs worse on non-factual and deceptive content, while prompting Large Language Models for the same task is less sensitive to these properties.
Authors: Hyeongdon Moon, Richard Davis, Seyed Parsa Neshaei, Pierre Dillenbourg
Abstract: Knowledge tracing models have enabled a range of intelligent tutoring systems to provide feedback to students. However, existing methods for knowledge tracing in learning sciences are predominantly reliant on statistical data and instructor-defined knowledge components, making it challenging to integrate AI-generated educational content with traditional established methods. We propose a method for automatically extracting knowledge components from educational content using instruction-tuned large multimodal models. We validate this approach by comprehensively evaluating it against knowledge tracing benchmarks in five domains. Our results indicate that the automatically extracted knowledge components can effectively replace human-tagged labels, offering a promising direction for enhancing intelligent tutoring systems in limited-data scenarios, achieving more explainable assessments in educational settings, and laying the groundwork for automated assessment.
Authors: Luohe Shi, Yao Yao, Zuchao Li, Lefei Zhang, Hai Zhao
Abstract: Large language models (LLMs) have rapidly advanced and demonstrated impressive capabilities. In-Context Learning (ICL) and Parameter-Efficient Fine-Tuning (PEFT) are currently two mainstream methods for augmenting LLMs to downstream tasks. ICL typically constructs a few-shot learning scenario, either manually or by setting up a Retrieval-Augmented Generation (RAG) system, helping models quickly grasp domain knowledge or question-answering patterns without changing model parameters. However, this approach involves trade-offs, such as slower inference speed and increased space occupancy. PEFT assists the model in adapting to tasks through minimal parameter modifications, but the training process still demands high hardware requirements, even with a small number of parameters involved. To address these challenges, we propose Reference Trustable Decoding (RTD), a paradigm that allows models to quickly adapt to new tasks without fine-tuning, maintaining low inference costs. RTD constructs a reference datastore from the provided training examples and optimizes the LLM's final vocabulary distribution by flexibly selecting suitable references based on the input, resulting in more trustable responses and enabling the model to adapt to downstream tasks at a low cost. Experimental evaluations on various LLMs using different benchmarks demonstrate that RTD establishes a new paradigm for augmenting models to downstream tasks. Furthermore, our method exhibits strong orthogonality with traditional methods, allowing for concurrent usage.
Authors: Areeg Fahad Rasheed, M. Zarkoosh, Safa F. Abbas, Sana Sabah Al-Azzawi
Abstract: This paper addresses the challenge of classifying and assigning programming tasks to experts, a process that typically requires significant effort, time, and cost. To tackle this issue, a novel dataset containing a total of 4,112 programming tasks was created by extracting tasks from various websites. Web scraping techniques were employed to collect this dataset of programming problems systematically. Specific HTML tags were tracked to extract key elements of each issue, including the title, problem description, input-output, examples, problem class, and complexity score. Examples from the dataset are provided in the appendix to illustrate the variety and complexity of tasks included. The dataset's effectiveness has been evaluated and benchmarked using two approaches; the first approach involved fine-tuning the FLAN-T5 small model on the dataset, while the second approach used in-context learning (ICL) with the GPT-4o mini. The performance was assessed using standard metrics: accuracy, recall, precision, and F1-score. The results indicated that in-context learning with GPT-4o-mini outperformed the FLAN-T5 model.
Authors: Jesujoba O. Alabi, Xuechen Liu, Dietrich Klakow, Junichi Yamagishi
Abstract: In this work, we present AfriHuBERT, an extension of mHuBERT-147, a state-of-the-art (SOTA) and compact self-supervised learning (SSL) model, originally pretrained on 147 languages. While mHuBERT-147 was pretrained on 16 African languages, we expand this to cover 39 African languages through continued pretraining on 6,500+ hours of speech data aggregated from diverse sources, including 23 newly added languages. We evaluate AfriHuBERT on two key speech tasks: Language Identification (LID) and Automatic Speech Recognition (ASR) using FLEURS dataset. Our results show a +4% F1 score improvement on average for LID and a -1.2% average Word Error Rate (WER) reduction for ASR. Further analysis shows that ASR models trained on AfriHuBERT exhibit improved cross-corpus generalization. Additionally, the analysis indicates that the FLEURS have data quality limitations that may affect their suitability for evaluating low-resource African languages, suggesting the need for better evaluation benchmarks for these languages.
Authors: Aditi Dutta, Susan Banducci, Chico Q. Camargo
Abstract: In recent years, several computational tools have been developed to detect and identify sexism, misogyny, and gender-based hate speech, especially on online platforms. Though these tools intend to draw on knowledge from both social science and computer science, little is known about the current state of research in quantifying online sexism or misogyny. Given the growing concern over the discrimination of women in online spaces and the rise in interdisciplinary research on capturing the online manifestation of sexism and misogyny, a systematic literature review on the research practices and their measures is the need of the hour. We make three main contributions: (i) we present a semi-automated way to narrow down the search results in the different phases of selection stage in the PRISMA flowchart; (ii) we perform a systematic literature review of research papers that focus on the quantification and measurement of online gender-based hate speech, examining literature from computer science and the social sciences from 2012 to 2022; and (iii) we identify the opportunities and challenges for measuring gender-based online hate speech. Our findings from topic analysis suggest a disciplinary divide between the themes of research on sexism/misogyny. With evidence-based review, we summarise the different approaches used by the studies who have explored interdisciplinary approaches to bridge the knowledge gap. Coupled with both the existing literature on social science theories and computational modeling, we provide an analysis of the benefits and shortcomings of the methodologies used. Lastly, we discuss the challenges and opportunities for future research dedicated to measuring online sexism and misogyny.
Authors: David Castillo-Bolado, Joseph Davidson, Finlay Gray, Marek Rosa
Abstract: We introduce a dynamic benchmarking system for conversational agents that evaluates their performance through a single, simulated, and lengthy user$\leftrightarrow$agent interaction. The interaction is a conversation between the user and agent, where multiple tasks are introduced and then undertaken concurrently. We context switch regularly to interleave the tasks, which constructs a realistic testing scenario in which we assess the Long-Term Memory, Continual Learning, and Information Integration capabilities of the agents. Results from both proprietary and open-source Large-Language Models show that LLMs in general perform well on single-task interactions, but they struggle on the same tasks when they are interleaved. Notably, short-context LLMs supplemented with an LTM system perform as well as or better than those with larger contexts. Our benchmark suggests that there are other challenges for LLMs responding to more natural interactions that contemporary benchmarks have heretofore not been able to capture.
Authors: Huachuan Qiu, Lizhi Ma, Zhenzhong Lan
Abstract: As awareness of mental health issues grows, online counseling support services are becoming increasingly prevalent worldwide. Detecting whether users express suicidal ideation in text-based counseling services is crucial for identifying and prioritizing at-risk individuals. However, the lack of domain-specific systems to facilitate fine-grained suicide detection and corresponding risk assessment in online counseling poses a significant challenge for automated crisis intervention aimed at suicide prevention. In this paper, we propose PsyGUARD, an automated system for detecting suicide ideation and assessing risk in psychological counseling. To achieve this, we first develop a detailed taxonomy for detecting suicide ideation based on foundational theories. We then curate a large-scale, high-quality dataset called PsySUICIDE for suicide detection. To evaluate the capabilities of automated systems in fine-grained suicide detection, we establish a range of baselines. Subsequently, to assist automated services in providing safe, helpful, and tailored responses for further assessment, we propose to build a suite of risk assessment frameworks. Our study not only provides an insightful analysis of the effectiveness of automated risk assessment systems based on fine-grained suicide detection but also highlights their potential to improve mental health services on online counseling platforms. Code, data, and models are available at https://github.com/qiuhuachuan/PsyGUARD.
Authors: Abdullah Barayan, Jose Camacho-Collados, Fernando Alva-Manchego
Abstract: Readability-controlled text simplification (RCTS) rewrites texts to lower readability levels while preserving their meaning. RCTS models often depend on parallel corpora with readability annotations on both source and target sides. Such datasets are scarce and difficult to curate, especially at the sentence level. To reduce reliance on parallel data, we explore using instruction-tuned large language models for zero-shot RCTS. Through automatic and manual evaluations, we examine: (1) how different types of contextual information affect a model's ability to generate sentences with the desired readability, and (2) the trade-off between achieving target readability and preserving meaning. Results show that all tested models struggle to simplify sentences (especially to the lowest levels) due to models' limitations and characteristics of the source sentences that impede adequate rewriting. Our experiments also highlight the need for better automatic evaluation metrics tailored to RCTS, as standard ones often misinterpret common simplification operations, and inaccurately assess readability and meaning preservation.
Authors: Haitao Li, You Chen, Qingyao Ai, Yueyue Wu, Ruizhe Zhang, Yiqun Liu
Abstract: Large language models (LLMs) have made significant progress in natural language processing tasks and demonstrate considerable potential in the legal domain. However, legal applications demand high standards of accuracy, reliability, and fairness. Applying existing LLMs to legal systems without careful evaluation of their potential and limitations could pose significant risks in legal practice. To this end, we introduce a standardized comprehensive Chinese legal benchmark LexEval. This benchmark is notable in the following three aspects: (1) Ability Modeling: We propose a new taxonomy of legal cognitive abilities to organize different tasks. (2) Scale: To our knowledge, LexEval is currently the largest Chinese legal evaluation dataset, comprising 23 tasks and 14,150 questions. (3) Data: we utilize formatted existing datasets, exam datasets and newly annotated datasets by legal experts to comprehensively evaluate the various capabilities of LLMs. LexEval not only focuses on the ability of LLMs to apply fundamental legal knowledge but also dedicates efforts to examining the ethical issues involved in their application. We evaluated 38 open-source and commercial LLMs and obtained some interesting findings. The experiments and findings offer valuable insights into the challenges and potential solutions for developing Chinese legal systems and LLM evaluation pipelines. The LexEval dataset and leaderboard are publicly available at \url{https://github.com/CSHaitao/LexEval} and will be continuously updated.
Authors: Laur\`ene Vaugrante, Mathias Niepert, Thilo Hagendorff
Abstract: In an era where large language models (LLMs) are increasingly integrated into a wide range of everyday applications, research into these models' behavior has surged. However, due to the novelty of the field, clear methodological guidelines are lacking. This raises concerns about the replicability and generalizability of insights gained from research on LLM behavior. In this study, we discuss the potential risk of a replication crisis and support our concerns with a series of replication experiments focused on prompt engineering techniques purported to influence reasoning abilities in LLMs. We tested GPT-3.5, GPT-4o, Gemini 1.5 Pro, Claude 3 Opus, Llama 3-8B, and Llama 3-70B, on the chain-of-thought, EmotionPrompting, ExpertPrompting, Sandbagging, as well as Re-Reading prompt engineering techniques, using manually double-checked subsets of reasoning benchmarks including CommonsenseQA, CRT, NumGLUE, ScienceQA, and StrategyQA. Our findings reveal a general lack of statistically significant differences across nearly all techniques tested, highlighting, among others, several methodological weaknesses in previous research. We propose a forward-looking approach that includes developing robust methodologies for evaluating LLMs, establishing sound benchmarks, and designing rigorous experimental frameworks to ensure accurate and reliable assessments of model outputs.
Authors: Linus Tze En Foo, Lynnette Hui Xian Ng
Abstract: Singlish, or formally Colloquial Singapore English, is an English-based creole language originating from the SouthEast Asian country Singapore. The language contains influences from Sinitic languages such as Chinese dialects, Malay, Tamil and so forth. A fundamental task to understanding Singlish is to first understand the pragmatic functions of its discourse particles, upon which Singlish relies heavily to convey meaning. This work offers a preliminary effort to disentangle the Singlish discourse particles (lah, meh and hor) with task-driven representation learning. After disentanglement, we cluster these discourse particles to differentiate their pragmatic functions, and perform Singlish-to-English machine translation. Our work provides a computational method to understanding Singlish discourse particles, and opens avenues towards a deeper comprehension of the language and its usage.
Authors: Tomilov A. A., Gromova A. Y., Svischev A. N
Abstract: In this paper we propose a word-wise intonation model for Russian language and show how it can be generalized for other languages. The proposed model is suitable for automatic data markup and its extended application to text-to-speech systems. It can also be implemented for an intonation contour modeling by using rule-based algorithms or by predicting contours with language models. The key idea is a partial elimination of the variability connected with different placements of a stressed syllable in a word. It is achieved with simultaneous applying of pitch simplification with a dynamic time warping clustering. The proposed model could be used as a tool for intonation research or as a backbone for prosody description in text-to-speech systems. As the advantage of the model, we show its relations with the existing intonation systems as well as the possibility of using language models for prosody prediction. Finally, we demonstrate some practical evidence of the system robustness to parameter variations.
Authors: Shan Chen, Mingye Gao, Kuleen Sasse, Thomas Hartvigsen, Brian Anthony, Lizhou Fan, Hugo Aerts, Jack Gallifant, Danielle Bitterman
Abstract: Background: Large language models (LLMs) are trained to follow directions, but this introduces a vulnerability to blindly comply with user requests even if they generate wrong information. In medicine, this could accelerate the generation of misinformation that impacts human well-being. Objectives/Methods: We analyzed compliance to requests to generate misleading content about medications in settings where models know the request is illogical. We investigated whether in-context directions and instruction-tuning of LLMs to prioritize logical reasoning over compliance reduced misinformation risk. Results: While all frontier LLMs complied with misinformation requests, both prompt-based and parameter-based approaches can improve the detection of logic flaws in requests and prevent the dissemination of medical misinformation. Conclusion: Shifting LLMs to prioritize logic over compliance could reduce risks of exploitation for medical misinformation.
Authors: Connor Baumler, Hal Daum\'e III
Abstract: AI-based systems such as language models can replicate and amplify social biases reflected in their training data. Among other questionable behavior, this can lead to LM-generated text--and text suggestions--that contain normatively inappropriate stereotypical associations. In this paper, we consider the question of how "debiasing" a language model impacts stories that people write using that language model in a predictive text scenario. We find that (n=414), in certain scenarios, language model suggestions that align with common social stereotypes are more likely to be accepted by human authors. Conversely, although anti-stereotypical language model suggestions sometimes lead to an increased rate of anti-stereotypical stories, this influence is far from sufficient to lead to "fully debiased" stories.
Authors: Runze Xia, Congchi Yin, Piji Li
Abstract: The human visual system is capable of processing continuous streams of visual information, but how the brain encodes and retrieves recent visual memories during continuous visual processing remains unexplored. This study investigates the capacity of working memory to retain past information under continuous visual stimuli. And then we propose a new task Memory Disentangling, which aims to extract and decode past information from fMRI signals. To address the issue of interference from past memory information, we design a disentangled contrastive learning method inspired by the phenomenon of proactive interference. This method separates the information between adjacent fMRI signals into current and past components and decodes them into image descriptions. Experimental results demonstrate that this method effectively disentangles the information within fMRI signals. This research could advance brain-computer interfaces and mitigate the problem of low temporal resolution in fMRI.
Authors: Fan Yuan, Chi Qin, Xiaogang Xu, Piji Li
Abstract: Large Vision-Language Models (LVLMs) have shown remarkable performance on many visual-language tasks. However, these models still suffer from multimodal hallucination, which means the generation of objects or content that violates the images. Many existing work detects hallucination by directly judging whether an object exists in an image, overlooking the association between the object and semantics. To address this issue, we propose Hierarchical Feedback Learning with Vision-enhanced Penalty Decoding (HELPD). This framework incorporates hallucination feedback at both object and sentence semantic levels. Remarkably, even with a marginal degree of training, this approach can alleviate over 15% of hallucination. Simultaneously, HELPD penalizes the output logits according to the image attention window to avoid being overly affected by generated text. HELPD can be seamlessly integrated with any LVLMs. Our experiments demonstrate that the proposed framework yields favorable results across multiple hallucination benchmarks. It effectively mitigates hallucination for different LVLMs and concurrently improves their text generation quality.
Authors: Zhengren Wang, Qinhan Yu, Shida Wei, Zhiyu Li, Feiyu Xiong, Xiaoxing Wang, Simin Niu, Hao Liang, Wentao Zhang
Abstract: Modern QA systems entail retrieval-augmented generation (RAG) for accurate and trustworthy responses. However, the inherent gap between user queries and relevant documents hinders precise matching. Motivated by our conical distribution hypothesis, which posits that potential queries and documents form a cone-like structure in the embedding space, we introduce QAEncoder, a training-free approach to bridge this gap. Specifically, QAEncoder estimates the expectation of potential queries in the embedding space as a robust surrogate for the document embedding, and attaches document fingerprints to effectively distinguish these embeddings. Extensive experiments on fourteen embedding models across six languages and eight datasets validate QAEncoder's alignment capability, which offers a plug-and-play solution that seamlessly integrates with existing RAG architectures and training-based methods.
Authors: Xiaosong Yuan, Chen Shen, Shaotian Yan, Xiaofeng Zhang, Liang Xie, Wenxiao Wang, Renchu Guan, Ying Wang, Jieping Ye
Abstract: Zero-shot Chain-of-Thought (CoT) prompting emerges as a simple and effective strategy for enhancing the performance of large language models (LLMs) in real-world reasoning tasks. Nonetheless, the efficacy of a singular, task-level prompt uniformly applied across the whole of instances is inherently limited since one prompt cannot be a good partner for all, a more appropriate approach should consider the interaction between the prompt and each instance meticulously. This work introduces an instance-adaptive prompting algorithm as an alternative zero-shot CoT reasoning scheme by adaptively differentiating good and bad prompts. Concretely, we first employ analysis on LLMs through the lens of information flow to detect the mechanism under zero-shot CoT reasoning, in which we discover that information flows from question to prompt and question to rationale jointly influence the reasoning results most. We notice that a better zero-shot CoT reasoning needs the prompt to obtain semantic information from the question then the rationale aggregates sufficient information from the question directly and via the prompt indirectly. On the contrary, lacking any of those would probably lead to a bad one. Stem from that, we further propose an instance-adaptive prompting strategy (IAP) for zero-shot CoT reasoning. Experiments conducted with LLaMA-2, LLaMA-3, and Qwen on math, logic, and commonsense reasoning tasks (e.g., GSM8K, MMLU, Causal Judgement) obtain consistent improvement, demonstrating that the instance-adaptive zero-shot CoT prompting performs better than other task-level methods with some curated prompts or sophisticated procedures, showing the significance of our findings in the zero-shot CoT reasoning mechanism.
Authors: Antonio Moreno-Sandoval, Leonardo Campillos-Llanos, Ana Garc\'ia-Serrano
Abstract: This work describes the language resources and models developed for automatic simplification of Spanish texts in three domains: Finance, Medicine and History studies. We created several corpora in each domain, annotation and simplification guidelines, a lexicon of technical and simplified medical terms, datasets used in shared tasks for the financial domain, and two simplification tools. The methodology, resources and companion publications are shared publicly on the web-site: https://clara-nlp.uned.es/.
Authors: Dung Ha Nguyen, Anh Thi Hoang Nguyen, Kiet Van Nguyen
Abstract: This study introduces an innovative automatic labeling framework to address the challenges of lexical normalization in social media texts for low-resource languages like Vietnamese. Social media data is rich and diverse, but the evolving and varied language used in these contexts makes manual labeling labor-intensive and expensive. To tackle these issues, we propose a framework that integrates semi-supervised learning with weak supervision techniques. This approach enhances the quality of training dataset and expands its size while minimizing manual labeling efforts. Our framework automatically labels raw data, converting non-standard vocabulary into standardized forms, thereby improving the accuracy and consistency of the training data. Experimental results demonstrate the effectiveness of our weak supervision framework in normalizing Vietnamese text, especially when utilizing Pre-trained Language Models. The proposed framework achieves an impressive F1-score of 82.72% and maintains vocabulary integrity with an accuracy of up to 99.22%. Additionally, it effectively handles undiacritized text under various conditions. This framework significantly enhances natural language normalization quality and improves the accuracy of various NLP tasks, leading to an average accuracy increase of 1-3%.
Authors: Vlad-Cristian Matei, Iulian-Marius T\u{a}iatu, R\u{a}zvan-Alexandru Sm\u{a}du, Dumitru-Clementin Cercel
Abstract: This paper highlights the significance of natural language processing (NLP) within artificial intelligence, underscoring its pivotal role in comprehending and modeling human language. Recent advancements in NLP, particularly in conversational bots, have garnered substantial attention and adoption among developers. This paper explores advanced methodologies for attaining smaller and more efficient NLP models. Specifically, we employ three key approaches: (1) training a Transformer-based neural network to detect offensive language, (2) employing data augmentation and knowledge distillation techniques to increase performance, and (3) incorporating multi-task learning with knowledge distillation and teacher annealing using diverse datasets to enhance efficiency. The culmination of these methods has yielded demonstrably improved outcomes.
Authors: Pablo Ortega, Jordi Luque, Luis Lamiable, Rodrigo L\'opez, Richard Benjamins
Abstract: Human language, while aimed at conveying meaning, inherently carries ambiguity. It poses challenges for speech and language processing, but also serves crucial communicative functions. Efficiently solve ambiguity is both a desired and a necessary characteristic. The lexical meaning of a word in context can be determined automatically by Word Sense Disambiguation (WSD) algorithms that rely on external knowledge often limited and biased toward English. When adapting content to other languages, automated translations are frequently inaccurate and a high degree of expert human validation is necessary to ensure both accuracy and understanding. The current study addresses previous limitations by introducing a new resource for Spanish WSD. It includes a sense inventory and a lexical dataset sourced from the Diccionario de la Lengua Espa\~nola which is maintained by the Real Academia Espa\~nola. We also review current resources for Spanish and report metrics on them by a state-of-the-art system.
Authors: Iker De la Iglesia, Iakes Goenaga, Johanna Ramirez-Romero, Jose Maria Villa-Gonzalez, Josu Goikoetxea, Ander Barrena
Abstract: Evaluating LLM-generated text has become a key challenge, especially in domain-specific contexts like the medical field. This work introduces a novel evaluation methodology for LLM-generated medical explanatory arguments, relying on Proxy Tasks and rankings to closely align results with human evaluation criteria, overcoming the biases typically seen in LLMs used as judges. We demonstrate that the proposed evaluators are robust against adversarial attacks, including the assessment of non-argumentative text. Additionally, the human-crafted arguments needed to train the evaluators are minimized to just one example per Proxy Task. By examining multiple LLM-generated arguments, we establish a methodology for determining whether a Proxy Task is suitable for evaluating LLM-generated medical explanatory arguments, requiring only five examples and two human experts.
Authors: Ralph P. Lano
Abstract: This paper presents the implementation of a self-replicating finite-state machine (FSM) and a self-replicating Turing Machine (TM) using bio-inspired mechanisms. Building on previous work that introduced self-replicating structures capable of sorting, copying, and reading information, this study demonstrates the computational power of these mechanisms by explicitly constructing a functioning FSM and TM. This study demonstrates the universality of the system by emulating the UTM(5,5) of Neary and Woods.
Authors: Haobo Li, Zhaowei Wang, Jiachen Wang, Alexis Kai Hon Lau, Huamin Qu
Abstract: Forecasting weather and climate events is crucial for making appropriate measures to mitigate environmental hazards and minimize associated losses. Previous research on environmental forecasting focuses on predicting numerical meteorological variables related to closed-set events rather than forecasting open-set events directly, which limits the comprehensiveness of event forecasting. We propose Weather and Climate Event Forecasting (WCEF), a new task that leverages meteorological raster data and textual event data to predict potential weather and climate events. However, due to difficulties in aligning multimodal data and the lack of sufficient supervised datasets, this task is challenging to accomplish. Therefore, we first propose a framework to align historical meteorological data with past weather and climate events using the large language model (LLM). In this framework, we construct a knowledge graph by using LLM to extract information about weather and climate events from a corpus of over 41k highly environment-focused news articles. Subsequently, we mapped these events with meteorological raster data, creating a supervised dataset, which is the largest and most novel for LLM tuning on the WCEF task. Finally, we introduced our aligned models, CLLMate (LLM for climate), a multimodal LLM to forecast weather and climate events using meteorological raster data. In evaluating CLLMate, we conducted extensive experiments. The results indicate that CLLMate surpasses both the baselines and other multimodal LLMs, showcasing the potential of utilizing LLM to align weather and climate events with meteorological data and highlighting the promising future for research on the WCEF task.
Authors: Diogo Gl\'oria-Silva, David Semedo, Jo\~ao Magalh\~aes
Abstract: Guiding users through complex procedural plans is an inherently multimodal task in which having visually illustrated plan steps is crucial to deliver an effective plan guidance. However, existing works on plan-following language models (LMs) often are not capable of multimodal input and output. In this work, we present MM-PlanLLM, the first multimodal LLM designed to assist users in executing instructional tasks by leveraging both textual plans and visual information. Specifically, we bring cross-modality through two key tasks: Conversational Video Moment Retrieval, where the model retrieves relevant step-video segments based on user queries, and Visually-Informed Step Generation, where the model generates the next step in a plan, conditioned on an image of the user's current progress. MM-PlanLLM is trained using a novel multitask-multistage approach, designed to gradually expose the model to multimodal instructional-plans semantic layers, achieving strong performance on both multimodal and textual dialogue in a plan-grounded setting. Furthermore, we show that the model delivers cross-modal temporal and plan-structure representations aligned between textual plan steps and instructional video moments.
Authors: In Gim, Caihua Li, Lin Zhong
Abstract: Our work tackles the challenge of securing user inputs in cloud-based large language model (LLM) services while ensuring output consistency, model confidentiality, and compute efficiency. We introduce Secure Multi-party Decoding (SMD), which leverages confidential computing to confine user prompts to a trusted execution environment, namely a confidential virtual machine (CVM), while allowing service providers to generate tokens efficiently. We also introduce a novel cryptographic method, Prompt Obfuscation (PO), to ensure robustness against reconstruction attacks on SMD. We demonstrate that our approach preserves both prompt confidentiality and LLM serving efficiency. Our solution can enable privacy-preserving cloud LLM services that handle sensitive prompts, such as clinical records, financial data, and personal information.
Authors: Michael S. Yao, Allison Chae, Charles E. Kahn Jr., Walter R. Witschey, James C. Gee, Hersh Sagreiya, Osbert Bastani
Abstract: Diagnostic imaging studies are an increasingly important component of the workup and management of acutely presenting patients. However, ordering appropriate imaging studies according to evidence-based medical guidelines is a challenging task with a high degree of variability between healthcare providers. To address this issue, recent work has investigated if generative AI and large language models can be leveraged to help clinicians order relevant imaging studies for patients. However, it is challenging to ensure that these tools are correctly aligned with medical guidelines, such as the American College of Radiology's Appropriateness Criteria (ACR AC). In this study, we introduce a framework to intelligently leverage language models by recommending imaging studies for patient cases that are aligned with evidence-based guidelines. We make available a novel dataset of patient "one-liner" scenarios to power our experiments, and optimize state-of-the-art language models to achieve an accuracy on par with clinicians in image ordering. Finally, we demonstrate that our language model-based pipeline can be used as intelligent assistants by clinicians to support image ordering workflows and improve the accuracy of imaging study ordering according to the ACR AC. Our work demonstrates and validates a strategy to leverage AI-based software to improve trustworthy clinical decision making in alignment with expert evidence-based guidelines.
Authors: Christine de Kock
Abstract: Group interactions take place within a particular socio-temporal context, which should be taken into account when modelling communities. We propose a method for jointly modelling community structure and language over time, and apply it in the context of extremist anti-women online groups (collectively known as the manosphere). Our model derives temporally grounded embeddings for words and users, which evolve over the training window. We show that this approach outperforms prior models which lacked one of these components (i.e. not incorporating social structure, or using static word embeddings). Using these embeddings, we investigate the evolution of users and words within these communities in three ways: (i) we model a user as a sequence of embeddings and forecast their affinity groups beyond the training window, (ii) we illustrate how word evolution is useful in the context of temporal events, and (iii) we characterise the propensity for violent language within subgroups of the manosphere.
Authors: Kazuki Matsuda, Yuiga Wada, Komei Sugiura
Abstract: In this work, we address the challenge of developing automatic evaluation metrics for image captioning, with a particular focus on robustness against hallucinations. Existing metrics are often inadequate for handling hallucinations, primarily due to their limited ability to compare candidate captions with multifaceted reference captions. To address this shortcoming, we propose DENEB, a novel supervised automatic evaluation metric specifically robust against hallucinations. DENEB incorporates the Sim-Vec Transformer, a mechanism that processes multiple references simultaneously, thereby efficiently capturing the similarity between an image, a candidate caption, and reference captions. To train DENEB, we construct the diverse and balanced Nebula dataset comprising 32,978 images, paired with human judgments provided by 805 annotators. We demonstrated that DENEB achieves state-of-the-art performance among existing LLM-free metrics on the FOIL, Composite, Flickr8K-Expert, Flickr8K-CF, Nebula, and PASCAL-50S datasets, validating its effectiveness and robustness against hallucinations.
Authors: Jiuding Duan, Jiyi Li, Yukino Baba, Hisashi Kashima
Abstract: Intransitivity is a critical issue in pairwise preference modeling. It refers to the intransitive pairwise preferences between a group of players or objects that potentially form a cyclic preference chain and has been long discussed in social choice theory in the context of the dominance relationship. However, such multifaceted intransitivity between players and the corresponding player representations in high dimensions is difficult to capture. In this paper, we propose a probabilistic model that jointly learns each player's d-dimensional representation (d>1) and a dataset-specific metric space that systematically captures the distance metric in Rd over the embedding space. Interestingly, by imposing additional constraints in the metric space, our proposed model degenerates to former models used in intransitive representation learning. Moreover, we present an extensive quantitative investigation of the vast existence of intransitive relationships between objects in various real-world benchmark datasets. To our knowledge, this investigation is the first of this type. The predictive performance of our proposed method on different real-world datasets, including social choice, election, and online game datasets, shows that our proposed method outperforms several competing methods in terms of prediction accuracy.
Authors: Chenxi Wang, Zongfang Liu, Dequan Yang, Xiuying Chen
Abstract: The impact of social media on critical issues such as echo chambers needs to be addressed, as these phenomena can have disruptive consequences for our society. Traditional research often oversimplifies emotional tendencies and opinion evolution into numbers and formulas, neglecting that news and communication are conveyed through text, which limits these approaches. Hence, in this work, we propose an LLM-based simulation for the social opinion network to evaluate and counter polarization phenomena. We first construct three typical network structures to simulate different characteristics of social interactions. Then, agents interact based on recommendation algorithms and update their strategies through reasoning and analysis. By comparing these interactions with the classic Bounded Confidence Model (BCM), the Friedkin Johnsen (FJ) model, and using echo chamber-related indices, we demonstrate the effectiveness of our framework in simulating opinion dynamics and reproducing phenomena such as opinion polarization and echo chambers. We propose two mitigation methods, active and passive nudges, that can help reduce echo chambers, specifically within language-based simulations. We hope our work will offer valuable insights and guidance for social polarization mitigation.
Authors: Zongbo Han, Jialong Yang, Junfan Li, Qinghua Hu, Qianli Xu, Mike Zheng Shou, Changqing Zhang
Abstract: Vision-language foundation models (e.g., CLIP) have shown remarkable performance across a wide range of tasks. However, deploying these models may be unreliable when significant distribution gaps exist between the training and test data. The training-free test-time dynamic adapter (TDA) is a promising approach to address this issue by storing representative test samples to guide the classification of subsequent ones. However, TDA only naively maintains a limited number of reference samples in the cache, leading to severe test-time catastrophic forgetting when the cache is updated by dropping samples. In this paper, we propose a simple yet effective method for DistributiOnal Test-time Adaptation (Dota). Instead of naively memorizing representative test samples, Dota continually estimates the distributions of test samples, allowing the model to continually adapt to the deployment environment. The test-time posterior probabilities are then computed using the estimated distributions based on Bayes' theorem for adaptation purposes. To further enhance the adaptability on the uncertain samples, we introduce a new human-in-the-loop paradigm which identifies uncertain samples, collects human-feedback, and incorporates it into the Dota framework. Extensive experiments validate that Dota enables CLIP to continually learn, resulting in a significant improvement compared to current state-of-the-art methods.
Authors: Shivani Kapania, William Agnew, Motahhare Eslami, Hoda Heidari, Sarah Fox
Abstract: The recent excitement around generative models has sparked a wave of proposals suggesting the replacement of human participation and labor in research and development--e.g., through surveys, experiments, and interviews--with synthetic research data generated by large language models (LLMs). We conducted interviews with 19 qualitative researchers to understand their perspectives on this paradigm shift. Initially skeptical, researchers were surprised to see similar narratives emerge in the LLM-generated data when using the interview probe. However, over several conversational turns, they went on to identify fundamental limitations, such as how LLMs foreclose participants' consent and agency, produce responses lacking in palpability and contextual depth, and risk delegitimizing qualitative research methods. We argue that the use of LLMs as proxies for participants enacts the surrogate effect, raising ethical and epistemological concerns that extend beyond the technical limitations of current models to the core of whether LLMs fit within qualitative ways of knowing.
Authors: Yi Wu, Zikang Xiong, Yiran Hu, Shreyash S. Iyengar, Nan Jiang, Aniket Bera, Lin Tan, Suresh Jagannathan
Abstract: Despite significant advancements in large language models (LLMs) that enhance robot agents' understanding and execution of natural language (NL) commands, ensuring the agents adhere to user-specified constraints remains challenging, particularly for complex commands and long-horizon tasks. To address this challenge, we present three key insights, equivalence voting, constrained decoding, and domain-specific fine-tuning, which significantly enhance LLM planners' capability in handling complex tasks. Equivalence voting ensures consistency by generating and sampling multiple Linear Temporal Logic (LTL) formulas from NL commands, grouping equivalent LTL formulas, and selecting the majority group of formulas as the final LTL formula. Constrained decoding then uses the generated LTL formula to enforce the autoregressive inference of plans, ensuring the generated plans conform to the LTL. Domain-specific fine-tuning customizes LLMs to produce safe and efficient plans within specific task domains. Our approach, Safe Efficient LLM Planner (SELP), combines these insights to create LLM planners to generate plans adhering to user commands with high confidence. We demonstrate the effectiveness and generalizability of SELP across different robot agents and tasks, including drone navigation and robot manipulation. For drone navigation tasks, SELP outperforms state-of-the-art planners by 10.8% in safety rate (i.e., finishing tasks conforming to NL commands) and by 19.8% in plan efficiency. For robot manipulation tasks, SELP achieves 20.4% improvement in safety rate. Our datasets for evaluating NL-to-LTL and robot task planning will be released in github.com/lt-asset/selp.
Authors: Taha Koleilat, Hojat Asgariandehkordi, Hassan Rivaz, Yiming Xiao
Abstract: Segmentation of anatomical structures and pathological regions in medical images is essential for modern clinical diagnosis, disease research, and treatment planning. While significant advancements have been made in deep learning-based segmentation techniques, many of these methods still suffer from limitations in data efficiency, generalizability, and interactivity. As a result, developing precise segmentation methods that require fewer labeled datasets remains a critical challenge in medical image analysis. Recently, the introduction of foundation models like CLIP and Segment-Anything-Model (SAM), with robust cross-domain representations, has paved the way for interactive and universal image segmentation. However, further exploration of these models for data-efficient segmentation in medical imaging is still needed and highly relevant. In this paper, we introduce MedCLIP-SAMv2, a novel framework that integrates the CLIP and SAM models to perform segmentation on clinical scans using text prompts, in both zero-shot and weakly supervised settings. Our approach includes fine-tuning the BiomedCLIP model with a new Decoupled Hard Negative Noise Contrastive Estimation (DHN-NCE) loss, and leveraging the Multi-modal Information Bottleneck (M2IB) to create visual prompts for generating segmentation masks from SAM in the zero-shot setting. We also investigate using zero-shot segmentation labels within a weakly supervised paradigm to enhance segmentation quality further. Extensive testing across four diverse segmentation tasks and medical imaging modalities (breast tumor ultrasound, brain tumor MRI, lung X-ray, and lung CT) demonstrates the high accuracy of our proposed framework. Our code is available at https://github.com/HealthX-Lab/MedCLIP-SAMv2.
Authors: Xiao Wang, Jianlong Wu, Zijia Lin, Fuzheng Zhang, Di Zhang, Liqiang Nie
Abstract: Recently, video-language understanding has achieved great success through large-scale pre-training. However, data scarcity remains a prevailing challenge. This study quantitatively reveals an "impossible trinity" among data quantity, diversity, and quality in pre-training datasets. Recent efforts seek to refine large-scale, diverse ASR datasets compromised by low quality through synthetic annotations. These methods successfully leverage useful information in multimodal video content (frames, tags, ASR transcripts, etc.) to refine the original annotations. Nevertheless, they struggle to mitigate noise within synthetic annotations and lack scalability as the dataset size expands. To address these issues, we introduce the Video DataFlywheel framework, which iteratively refines video annotations with improved noise control methods. For iterative refinement, we first leverage a video-language model to generate synthetic annotations, resulting in a refined dataset. Then, we pre-train on it and fine-tune on human refinement examples for a stronger model. These processes are repeated for continuous improvement. For noise control, we present AdaTaiLr, a novel noise control method that requires weaker assumptions on noise distribution, thereby proving more effective in large datasets with theoretical guarantees. The combination of iterative refinement and AdaTaiLr can achieve better scalability in video-language understanding. Extensive experiments show that our framework outperforms existing data refinement baselines, delivering a 3% performance boost and improving dataset quality with minimal diversity loss. Furthermore, our refined dataset facilitates significant improvements in various video-language understanding tasks, including video question answering and text-video retrieval.
Authors: Chen Chen, Xiaolou Li, Zehua Liu, Lantian Li, Dong Wang
Abstract: In the field of spoken language processing, audio-visual speech processing is receiving increasing research attention. Key components of this research include tasks such as lip reading, audio-visual speech recognition, and visual-to-speech synthesis. Although significant success has been achieved, theoretical analysis is still insufficient for audio-visual tasks. This paper presents a quantitative analysis based on information theory, focusing on information intersection between different modalities. Our results show that this analysis is valuable for understanding the difficulties of audio-visual processing tasks as well as the benefits that could be obtained by modality integration.
Authors: Jinyi Mi, Xiaohan Shi, Ding Ma, Jiajun He, Takuya Fujimura, Tomoki Toda
Abstract: Developing a robust speech emotion recognition (SER) system in noisy conditions faces challenges posed by different noise properties. Most previous studies have not considered the impact of human speech noise, thus limiting the application scope of SER. In this paper, we propose a novel two-stage framework for the problem by cascading target speaker extraction (TSE) method and SER. We first train a TSE model to extract the speech of target speaker from a mixture. Then, in the second stage, we utilize the extracted speech for SER training. Additionally, we explore a joint training of TSE and SER models in the second stage. Our developed system achieves a 14.33% improvement in unweighted accuracy (UA) compared to a baseline without using TSE method, demonstrating the effectiveness of our framework in mitigating the impact of human speech noise. Moreover, we conduct experiments considering speaker gender, showing that our framework performs particularly well in different-gender mixture.
Authors: Ruizhe Shi, Runlong Zhou, Simon S. Du
Abstract: Direct Preference Optimization (DPO) has emerged as a stable, scalable, and efficient solution for language model alignment. Despite its empirical success, the $\textit{optimization}$ properties, particularly the impact of samplers on its convergence rates, remain underexplored. In this paper, we provide a rigorous analysis of DPO's $\textit{convergence rates}$ with different sampling strategies under the exact gradient setting, revealing a surprising separation: uniform sampling achieves $\textit{linear}$ convergence, while our proposed online sampler achieves $\textit{quadratic}$ convergence. We further adapt the sampler to practical settings by incorporating posterior distributions and $\textit{logit mixing}$, demonstrating significant improvements over previous approaches. On Safe-RLHF dataset, our method exhibits a $4.5$% improvement over vanilla DPO and a $3.0$% improvement over on-policy DPO; on Iterative-Prompt, our approach outperforms vanilla DPO, on-policy DPO, and Hybrid GSHF by over $4.2$%. Our results not only offer insights into the theoretical standing of DPO but also pave the way for potential algorithm designs in the future.
Authors: Defa Zhu, Hongzhi Huang, Zihao Huang, Yutao Zeng, Yunyao Mao, Banggu Wu, Qiyang Min, Xun Zhou
Abstract: We present hyper-connections, a simple yet effective method that can serve as an alternative to residual connections. This approach specifically addresses common drawbacks observed in residual connection variants, such as the seesaw effect between gradient vanishing and representation collapse. Theoretically, hyper-connections allow the network to adjust the strength of connections between features at different depths and dynamically rearrange layers. We conduct experiments focusing on the pre-training of large language models, including dense and sparse models, where hyper-connections show significant performance improvements over residual connections. Additional experiments conducted on vision tasks also demonstrate similar improvements. We anticipate that this method will be broadly applicable and beneficial across a wide range of AI problems.
Authors: Bikang Pan, Wei Huang, Ye Shi
Abstract: Integrating pretrained vision-language foundation models like CLIP into federated learning has attracted significant attention for enhancing generalization across diverse tasks. Typically, federated learning of vision-language models employs prompt learning to reduce communication and computational costs, i.e., prompt-based federated learning. However, there is limited theoretical analysis to understand the performance of prompt-based federated learning. In this work, we construct a theoretical analysis framework for prompt-based federated learning via feature learning theory. Specifically, we monitor the evolution of signal learning and noise memorization in prompt-based federated learning, demonstrating that performance can be assessed by the ratio of task-relevant to task-irrelevant coefficients. Furthermore, we draw an analogy between income and risk in portfolio optimization and the task-relevant and task-irrelevant terms in feature learning. Leveraging inspiration from portfolio optimization that combining two independent assets will maintain the income while reducing the risk, we introduce two prompts: global prompt and local prompt to construct a prompt portfolio to balance the generalization and personalization. Consequently, we showed the performance advantage of the prompt portfolio and derived the optimal mixing coefficient. These theoretical claims have been further supported by empirical experiments.
Authors: Youssef Hmamouche, Ismail Chihab, Lahoucine Kdouri, Amal El Fallah Seghrouchni
Abstract: Brain-related research topics in artificial intelligence have recently gained popularity, particularly due to the expansion of what multimodal architectures can do from computer vision to natural language processing. Our main goal in this work is to explore the possibilities and limitations of these architectures in spoken text decoding from non-invasive fMRI recordings. Contrary to vision and textual data, fMRI data represent a complex modality due to the variety of brain scanners, which implies (i) the variety of the recorded signal formats, (ii) the low resolution and noise of the raw signals, and (iii) the scarcity of pretrained models that can be leveraged as foundation models for generative learning. These points make the problem of the non-invasive decoding of text from fMRI recordings very challenging. In this paper, we propose and end-to-end multimodal LLM for decoding spoken text from fMRI signals. The proposed architecture is founded on (i) an encoder derived from a specific transformer incorporating an augmented embedding layer for the encoder and a better-adjusted attention mechanism than that present in the state of the art, and (ii) a frozen large language model adapted to align the embedding of the input text and the encoded embedding of brain activity to decode the output text. A benchmark in performed on a corpus consisting of a set of interactions human-human and human-robot interactions where fMRI and conversational signals are recorded synchronously. The obtained results are very promising, as our proposal outperforms the evaluated models, and is able to generate text capturing more accurate semantics present in the ground truth. The implementation code is provided in https://github.com/Hmamouche/brain_decode.
Authors: Rui Pan, Tuan Dung Nguyen, Hardik Arora, Alberto Accomazzi, Tirthankar Ghosal, Yuan-Sen Ting
Abstract: Continual pretraining of large language models on domain-specific data has been proposed to enhance performance on downstream tasks. In astronomy, the previous absence of astronomy-focused benchmarks has hindered objective evaluation of these specialized LLM models. Leveraging a recent initiative to curate high-quality astronomical MCQs, this study aims to quantitatively assess specialized LLMs in astronomy. We find that the previously released AstroLLaMA series, based on LLaMA-2-7B, underperforms compared to the base model. We demonstrate that this performance degradation can be partially mitigated by utilizing high-quality data for continual pretraining, such as summarized text from arXiv. Despite the observed catastrophic forgetting in smaller models, our results indicate that continual pretraining on the 70B model can yield significant improvements. However, the current supervised fine-tuning dataset still constrains the performance of instruct models. In conjunction with this study, we introduce a new set of models, AstroLLaMA-3-8B and AstroLLaMA-2-70B, building upon the previous AstroLLaMA series.
Authors: Atharva Naik, Marcus Alenius, Daniel Fried, Carolyn Rose
Abstract: The task of automated code review has recently gained a lot of attention from the machine learning community. However, current review comment evaluation metrics rely on comparisons with a human-written reference for a given code change (also called a diff), even though code review is a one-to-many problem like generation and summarization with many "valid reviews" for a diff. To tackle these issues we develop a CRScore - a reference-free metric to measure dimensions of review quality like conciseness, comprehensiveness, and relevance. We design CRScore to evaluate reviews in a way that is grounded in claims and potential issues detected in the code by LLMs and static analyzers. We demonstrate that CRScore can produce valid, fine-grained scores of review quality that have the greatest alignment with human judgment (0.54 Spearman correlation) and are more sensitive than reference-based metrics. We also release a corpus of 2.6k human-annotated review quality scores for machine-generated and GitHub review comments to support the development of automated metrics.
Authors: Johnathan Xie, Annie S. Chen, Yoonho Lee, Eric Mitchell, Chelsea Finn
Abstract: The effectiveness of large language models (LLMs) is not only measured by their ability to generate accurate outputs but also by their calibration-how well their confidence scores reflect the probability of their outputs being correct. While unsupervised pre-training has been shown to yield LLMs with well-calibrated conditional probabilities, recent studies have shown that after fine-tuning with reinforcement learning from human feedback (RLHF), the calibration of these models degrades significantly. In this work, we introduce Adaptive Temperature Scaling (ATS), a post-hoc calibration method that predicts a temperature scaling parameter for each token prediction. The predicted temperature values adapt based on token-level features and are fit over a standard supervised fine-tuning (SFT) dataset. The adaptive nature of ATS addresses the varying degrees of calibration shift that can occur after RLHF fine-tuning. ATS improves calibration by over 10-50% across three downstream natural language evaluation benchmarks compared to prior calibration methods and does not impede performance improvements from RLHF.
Authors: Ezra Karger, Houtan Bastani, Chen Yueh-Han, Zachary Jacobs, Danny Halawi, Fred Zhang, Philip E. Tetlock
Abstract: Forecasts of future events are essential inputs into informed decision-making. Machine learning (ML) systems have the potential to deliver forecasts at scale, but there is no framework for evaluating the accuracy of ML systems on a standardized set of forecasting questions. To address this gap, we introduce ForecastBench: a dynamic benchmark that evaluates the accuracy of ML systems on an automatically generated and regularly updated set of 1,000 forecasting questions. To avoid any possibility of data leakage, ForecastBench is comprised solely of questions about future events that have no known answer at the time of submission. We quantify the ability of current ML systems by collecting forecasts from expert (human) forecasters, the general public, and LLMs on a random subset of questions from the benchmark (N = 200). While LLMs have achieved super-human performance on many benchmarks, they perform less well here: expert forecasters outperform the top-performing LLM (p-values <= 0.01). We display system and human scores in a public leaderboard at www.forecastbench.org.
Authors: Shuhao Chen, Weisen Jiang, Baijiong Lin, James T. Kwok, Yu Zhang
Abstract: Recent works show that assembling multiple off-the-shelf large language models (LLMs) can harness their complementary abilities. To achieve this, routing is a promising method, which learns a router to select the most suitable LLM for each query. However, existing routing models are ineffective when multiple LLMs perform well for a query. To address this problem, in this paper, we propose a method called query-based Router by Dual Contrastive learning (RouterDC). The RouterDC model consists of an encoder and LLM embeddings, and we propose two contrastive learning losses to train the RouterDC model. Experimental results show that RouterDC is effective in assembling LLMs and largely outperforms individual top-performing LLMs as well as existing routing methods on both in-distribution (+2.76\%) and out-of-distribution (+1.90\%) tasks. Source code is available at https://github.com/shuhao02/RouterDC.
Authors: Johan Bjorck, Alon Benhaim, Vishrav Chaudhary, Furu Wei, Xia Song
Abstract: State-of-the-art LLMs are powered by scaling -- scaling model size, dataset size and cluster size. It is economically infeasible to extensively tune hyperparameter for the largest runs. Instead, approximately optimal hyperparameters must be inferred or \textit{transferred} from smaller experiments. Hyperparameter transfer across model sizes has been studied in Yang et al. However, hyperparameter transfer across dataset size -- or token horizon -- has not been studied yet. To remedy this we conduct a large scale empirical study on how optimal learning rate (LR) depends on token horizon in LLM training. We first demonstrate that the optimal LR changes significantly with token horizon -- longer training necessitates smaller LR. Secondly we demonstrate the the optimal LR follows a scaling law, and that the optimal LR for longer horizons can be accurately estimated from shorter horizons via our scaling laws. We also provide a rule-of-thumb for transferring LR across token horizons with zero overhead over current practices. Lastly we provide evidence that LLama-1 used too high LR, and estimate the performance hit from this. We thus argue that hyperparameter transfer across data size is an important and overlooked component of LLM training.
Authors: Qidong Liu, Xian Wu, Wanyu Wang, Yejing Wang, Yuanshao Zhu, Xiangyu Zhao, Feng Tian, Yefeng Zheng
Abstract: Sequential Recommender Systems (SRS) are extensively applied across various domains to predict users' next interaction by modeling their interaction sequences. However, these systems typically grapple with the long-tail problem, where they struggle to recommend items that are less popular. This challenge results in a decline in user discovery and reduced earnings for vendors, negatively impacting the system as a whole. Large Language Model (LLM) has the potential to understand the semantic connections between items, regardless of their popularity, positioning them as a viable solution to this dilemma. In our paper, we present LLMEmb, an innovative technique that harnesses LLM to create item embeddings that bolster the performance of SRS. To align the capabilities of general-purpose LLM with the needs of the recommendation domain, we introduce a method called Supervised Contrastive Fine-Tuning (SCFT). This method involves attribute-level data augmentation and a custom contrastive loss designed to tailor LLM for enhanced recommendation performance. Moreover, we highlight the necessity of incorporating collaborative filtering signals into LLM-generated embeddings and propose Recommendation Adaptation Training (RAT) for this purpose. RAT refines the embeddings to be optimally suited for SRS. The embeddings derived from LLMEmb can be easily integrated with any SRS model, showcasing its practical utility. Extensive experimentation on three real-world datasets has shown that LLMEmb significantly improves upon current methods when applied across different SRS models.
Authors: Ming Zhong, Aston Zhang, Xuewei Wang, Rui Hou, Wenhan Xiong, Chenguang Zhu, Zhengxing Chen, Liang Tan, Chloe Bi, Mike Lewis, Sravya Popuri, Sharan Narang, Melanie Kambadur, Dhruv Mahajan, Sergey Edunov, Jiawei Han, Laurens van der Maaten
Abstract: The development and evaluation of Large Language Models (LLMs) have largely focused on individual capabilities. However, this overlooks the intersection of multiple abilities across different types of expertise that are often required for real-world tasks, which we term cross capabilities. To systematically explore this concept, we first define seven core individual capabilities and then pair them to form seven common cross capabilities, each supported by a manually constructed taxonomy. Building on these definitions, we introduce CrossEval, a benchmark comprising 1,400 human-annotated prompts, with 100 prompts for each individual and cross capability. To ensure reliable evaluation, we involve expert annotators to assess 4,200 model responses, gathering 8,400 human ratings with detailed explanations to serve as reference examples. Our findings reveal that, in both static evaluations and attempts to enhance specific abilities, current LLMs consistently exhibit the "Law of the Weakest Link," where cross-capability performance is significantly constrained by the weakest component. Specifically, across 58 cross-capability scores from 17 models, 38 scores are lower than all individual capabilities, while 20 fall between strong and weak, but closer to the weaker ability. These results highlight the under-performance of LLMs in cross-capability tasks, making the identification and improvement of the weakest capabilities a critical priority for future research to optimize performance in complex, multi-dimensional scenarios.
Authors: Joshua Feinglass, Yezhou Yang
Abstract: Zero-shot inference, where pre-trained models perform tasks without specific training data, is an exciting emergent ability of large models like CLIP. Although there has been considerable exploration into enhancing zero-shot abilities in image captioning (IC) for popular datasets such as MSCOCO and Flickr8k, these approaches fall short with fine-grained datasets like CUB, FLO, UCM-Captions, and Sydney-Captions. These datasets require captions to discern between visually and semantically similar classes, focusing on detailed object parts and their attributes. To overcome this challenge, we introduce TRaining-Free Object-Part Enhancement (TROPE). TROPE enriches a base caption with additional object-part details using object detector proposals and Natural Language Processing techniques. It complements rather than alters the base caption, allowing seamless integration with other captioning methods and offering users enhanced flexibility. Our evaluations show that TROPE consistently boosts performance across all tested zero-shot IC approaches and achieves state-of-the-art results on fine-grained IC datasets.
Authors: Yabing Wang, Le Wang, Qiang Zhou, Zhibin Wang, Hao Li, Gang Hua, Wei Tang
Abstract: Cross-lingual cross-modal retrieval (CCR) aims to retrieve visually relevant content based on non-English queries, without relying on human-labeled cross-modal data pairs during training. One popular approach involves utilizing machine translation (MT) to create pseudo-parallel data pairs, establishing correspondence between visual and non-English textual data. However, aligning their representations poses challenges due to the significant semantic gap between vision and text, as well as the lower quality of non-English representations caused by pre-trained encoders and data noise. To overcome these challenges, we propose LECCR, a novel solution that incorporates the multi-modal large language model (MLLM) to improve the alignment between visual and non-English representations. Specifically, we first employ MLLM to generate detailed visual content descriptions and aggregate them into multi-view semantic slots that encapsulate different semantics. Then, we take these semantic slots as internal features and leverage them to interact with the visual features. By doing so, we enhance the semantic information within the visual features, narrowing the semantic gap between modalities and generating local visual semantics for subsequent multi-level matching. Additionally, to further enhance the alignment between visual and non-English features, we introduce softened matching under English guidance. This approach provides more comprehensive and reliable inter-modal correspondences between visual and non-English features. Extensive experiments on four CCR benchmarks, \ie Multi30K, MSCOCO, VATEX, and MSR-VTT-CN, demonstrate the effectiveness of our proposed method. Code: \url{https://github.com/LiJiaBei-7/leccr}.
Authors: Xinfeng Wang, Jin Cui, Fumiyo Fukumoto, Yoshimi Suzuki
Abstract: Large language models (LLMs) have demonstrated prominent reasoning capabilities in recommendation tasks by transforming them into text-generation tasks. % many NLP applications including However, existing approaches either disregard or ineffectively model the user--item high-order interactions. To this end, this paper presents an enhanced LLM-based recommender (ELMRec). We enhance whole-word embeddings to substantially enhance LLMs' interpretation of graph-constructed interactions for recommendations, without requiring graph pre-training. This finding may inspire endeavors to incorporate rich knowledge graphs into LLM-based recommenders via whole-word embedding. We also found that LLMs often recommend items based on users' earlier interactions rather than recent ones, and present a reranking solution. Our ELMRec outperforms state-of-the-art (SOTA) methods in both direct and sequential recommendations.
Authors: Oswald Zink, Yosuke Higuchi, Carlos Mullov, Alexander Waibel, Tetsunori Kobayashi
Abstract: Effective spoken dialog systems should facilitate natural interactions with quick and rhythmic timing, mirroring human communication patterns. To reduce response times, previous efforts have focused on minimizing the latency in automatic speech recognition (ASR) to optimize system efficiency. However, this approach requires waiting for ASR to complete processing until a speaker has finished speaking, which limits the time available for natural language processing (NLP) to formulate accurate responses. As humans, we continuously anticipate and prepare responses even while the other party is still speaking. This allows us to respond appropriately without missing the optimal time to speak. In this work, as a pioneering study toward a conversational system that simulates such human anticipatory behavior, we aim to realize a function that can predict the forthcoming words and estimate the time remaining until the end of an utterance (EOU), using the middle portion of an utterance. To achieve this, we propose a training strategy for an encoder-decoder-based ASR system, which involves masking future segments of an utterance and prompting the decoder to predict the words in the masked audio. Additionally, we develop a cross-attention-based algorithm that incorporates both acoustic and linguistic information to accurately detect the EOU. The experimental results demonstrate the proposed model's ability to predict upcoming words and estimate future EOU events up to 300ms prior to the actual EOU. Moreover, the proposed training strategy exhibits general improvements in ASR performance.
Authors: Qin Liu, Wenjie Mo, Terry Tong, Jiashu Xu, Fei Wang, Chaowei Xiao, Muhao Chen
Abstract: The advancement of Large Language Models (LLMs) has significantly impacted various domains, including Web search, healthcare, and software development. However, as these models scale, they become more vulnerable to cybersecurity risks, particularly backdoor attacks. By exploiting the potent memorization capacity of LLMs, adversaries can easily inject backdoors into LLMs by manipulating a small portion of training data, leading to malicious behaviors in downstream applications whenever the hidden backdoor is activated by the pre-defined triggers. Moreover, emerging learning paradigms like instruction tuning and reinforcement learning from human feedback (RLHF) exacerbate these risks as they rely heavily on crowdsourced data and human feedback, which are not fully controlled. In this paper, we present a comprehensive survey of emerging backdoor threats to LLMs that appear during LLM development or inference, and cover recent advancement in both defense and detection strategies for mitigating backdoor threats to LLMs. We also outline key challenges in addressing these threats, highlighting areas for future research.
Authors: Ke-Han Lu, Zhehuai Chen, Szu-Wei Fu, Chao-Han Huck Yang, Jagadeesh Balam, Boris Ginsburg, Yu-Chiang Frank Wang, Hung-yi Lee
Abstract: Recent end-to-end speech language models (SLMs) have expanded upon the capabilities of large language models (LLMs) by incorporating pre-trained speech models. However, these SLMs often undergo extensive speech instruction-tuning to bridge the gap between speech and text modalities. This requires significant annotation efforts and risks catastrophic forgetting of the original language capabilities. In this work, we present a simple yet effective automatic process for creating speech-text pair data that carefully injects speech paralinguistic understanding abilities into SLMs while preserving the inherent language capabilities of the text-based LLM. Our model demonstrates general capabilities for speech-related tasks without the need for speech instruction-tuning data, achieving impressive performance on Dynamic-SUPERB and AIR-Bench-Chat benchmarks. Furthermore, our model exhibits the ability to follow complex instructions derived from LLMs, such as specific output formatting and chain-of-thought reasoning. Our approach not only enhances the versatility and effectiveness of SLMs but also reduces reliance on extensive annotated datasets, paving the way for more efficient and capable speech understanding systems.
Authors: Frank Wawrzik, Matthias Plaue, Savan Vekariya, Christoph Grimm
Abstract: In this paper we propose a novel approach based on knowledge graphs to provide timely access to structured information, to enable actionable technology intelligence, and improve cyber-physical systems planning. Our framework encompasses a text mining process, which includes information retrieval, keyphrase extraction, semantic network creation, and topic map visualization. Following this data exploration process, we employ a selective knowledge graph construction (KGC) approach supported by an electronics and innovation ontology-backed pipeline for multi-objective decision-making with a focus on cyber-physical systems. We apply our methodology to the domain of automotive electrical systems to demonstrate the approach, which is scalable. Our results demonstrate that our construction process outperforms GraphGPT as well as our bi-LSTM and transformer REBEL with a pre-defined dataset by several times in terms of class recognition, relationship construction and correct "sublass of" categorization. Additionally, we outline reasoning applications and provide a comparison with Wikidata to show the differences and advantages of the approach.
Authors: Sheng Ouyang, Yulan Hu, Ge Chen, Yong Liu
Abstract: Large Language Models (LLMs) have achieved impressive results in processing text data, which has sparked interest in applying these models beyond textual data, such as graphs. In the field of graph learning, there is a growing interest in harnessing LLMs to comprehend and manipulate graph-structured data. Existing research predominantly focuses on graphs with rich textual features, such as knowledge graphs or text attribute graphs, leveraging LLMs' ability to process text but inadequately addressing graph structure. This work specifically aims to assess and enhance LLMs' abilities to comprehend and utilize the structural knowledge inherent in graph data itself, rather than focusing solely on graphs rich in textual content. To achieve this, we introduce the \textbf{G}raph \textbf{U}nderstanding for \textbf{N}atural Language \textbf{D}riven \textbf{A}nalytical \textbf{M}odel (\model). This model adapts LLMs to better understand and engage with the structure of graph data, enabling them to perform complex reasoning tasks by leveraging the graph's structure itself. Our experimental evaluations on graph reasoning benchmarks not only substantiate that \model~ outperforms the SOTA baselines for comparisons. But also reveals key factors affecting the graph reasoning capabilities of LLMs. Moreover, we provide a theoretical analysis illustrating how reasoning paths can enhance LLMs' reasoning capabilities.
Authors: Lei Yu, Virginie Do, Karen Hambardzumyan, Nicola Cancedda
Abstract: Large language models (LLMs) are vulnerable to adversarial attacks that can elicit harmful responses. Defending against such attacks remains challenging due to the opacity of jailbreaking mechanisms and the high computational cost of training LLMs robustly. We demonstrate that adversarial attacks share a universal mechanism for circumventing LLM safeguards that works by ablating a dimension in the residual stream embedding space called the refusal feature. We further show that the operation of refusal feature ablation (RFA) approximates the worst-case perturbation of offsetting model safety. Based on these findings, we propose Refusal Feature Adversarial Training (ReFAT), a novel algorithm that efficiently performs LLM adversarial training by simulating the effect of input-level attacks via RFA. Experiment results show that ReFAT significantly improves the robustness of three popular LLMs against a wide range of adversarial attacks, with considerably less computational overhead compared to existing adversarial training methods.
Authors: Zezhou Wang, Yaxin Du, Zhuzhong Qian, Siheng Chen
Abstract: Federated Domain-specific Instruction Tuning (FedDIT) leverages a few cross-client private data and server-side public data for instruction augmentation, enhancing model performance in specific domains. While the factors affecting FedDIT remain unclear and existing instruction augmentation methods mainly focus on the centralized setting without considering the distributed environment. Firstly, our experiments show that cross-client domain coverage, rather than data heterogeneity, drives model performance in FedDIT. Thus, we propose FedDCA, which maximizes domain coverage through greedy client center selection and retrieval-based augmentation. To reduce client-side computation, FedDCA$^*$ uses heterogeneous encoders with server-side feature alignment. Extensive experiments across four domains (code, medical, financial, and mathematical) validate the effectiveness of both methods. Additionally, we explore the privacy protection against memory extraction attacks with various amounts of public data and results show that there is no significant correlation between the amount of public data and the privacy-preserving capability. However, as the fine-tuning round increases, the risk of privacy leakage reduces or converges.
Authors: Zeyu Zhang, Quanyu Dai, Luyu Chen, Zeren Jiang, Rui Li, Jieming Zhu, Xu Chen, Yi Xie, Zhenhua Dong, Ji-Rong Wen
Abstract: LLM-based agents have been widely applied as personal assistants, capable of memorizing information from user messages and responding to personal queries. However, there still lacks an objective and automatic evaluation on their memory capability, largely due to the challenges in constructing reliable questions and answers (QAs) according to user messages. In this paper, we propose MemSim, a Bayesian simulator designed to automatically construct reliable QAs from generated user messages, simultaneously keeping their diversity and scalability. Specifically, we introduce the Bayesian Relation Network (BRNet) and a causal generation mechanism to mitigate the impact of LLM hallucinations on factual information, facilitating the automatic creation of an evaluation dataset. Based on MemSim, we generate a dataset in the daily-life scenario, named MemDaily, and conduct extensive experiments to assess the effectiveness of our approach. We also provide a benchmark for evaluating different memory mechanisms in LLM-based agents with the MemDaily dataset. To benefit the research community, we have released our project at https://github.com/nuster1128/MemSim.
Authors: Thomas P. Zollo, Andrew Wei Tung Siah, Naimeng Ye, Ang Li, Hongseok Namkoong
Abstract: As LLMs become capable of complex tasks, there is growing potential for personalized interactions tailored to the subtle and idiosyncratic preferences of the user. We present a public benchmark, PersonalLLM, focusing on adapting LLMs to provide maximal benefits for a particular user. Departing from existing alignment benchmarks that implicitly assume uniform preferences, we curate open-ended prompts paired with many high-quality answers over which users would be expected to display heterogeneous latent preferences. Instead of persona-prompting LLMs based on high-level attributes (e.g., user's race or response length), which yields homogeneous preferences relative to humans, we develop a method that can simulate a large user base with diverse preferences from a set of pre-trained reward models. Our dataset and generated personalities offer an innovative testbed for developing personalization algorithms that grapple with continual data sparsity--few relevant feedback from the particular user--by leveraging historical data from other (similar) users. We explore basic in-context learning and meta-learning baselines to illustrate the utility of PersonalLLM and highlight the need for future methodological development. Our dataset is available at https://huggingface.co/datasets/namkoong-lab/PersonalLLM
URLs: https://huggingface.co/datasets/namkoong-lab/PersonalLLM
Authors: Takafumi Moriya, Shota Horiguchi, Marc Delcroix, Ryo Masumura, Takanori Ashihara, Hiroshi Sato, Kohei Matsuura, Masato Mimura
Abstract: Extending the RNN Transducer (RNNT) to recognize multi-talker speech is essential for wider automatic speech recognition (ASR) applications. Multi-talker RNNT (MT-RNNT) aims to achieve recognition without relying on costly front-end source separation. MT-RNNT is conventionally implemented using architectures with multiple encoders or decoders, or by serializing all speakers' transcriptions into a single output stream. The first approach is computationally expensive, particularly due to the need for multiple encoder processing. In contrast, the second approach involves a complex label generation process, requiring accurate timestamps of all words spoken by all speakers in the mixture, obtained from an external ASR system. In this paper, we propose a novel alignment-free training scheme for the MT-RNNT (MT-RNNT-AFT) that adopts the standard RNNT architecture. The target labels are created by appending a prompt token corresponding to each speaker at the beginning of the transcription, reflecting the order of each speaker's appearance in the mixtures. Thus, MT-RNNT-AFT can be trained without relying on accurate alignments, and it can recognize all speakers' speech with just one round of encoder processing. Experiments show that MT-RNNT-AFT achieves performance comparable to that of the state-of-the-art alternatives, while greatly simplifying the training process.
Authors: Zhangcheng Qiang, Kerry Taylor
Abstract: Due to the dynamic nature of the semantic web, ontology version control is required to capture time-varying information, most importantly for widely-used ontologies. Despite the long-standing recognition of ontology versioning (OV) as a crucial component for efficient ontology management, the growing size of ontologies and accumulating errors caused by manual labour overwhelm current OV approaches. In this paper, we propose yet another approach to performing OV using existing ontology matching (OM) techniques and systems. We introduce a unified OM4OV pipeline. From an OM perspective, we reconstruct a new task formulation, performance measurement, and dataset construction for OV tasks. Reusing the prior alignment(s) from OM, we also propose a cross-reference mechanism to effectively reduce the matching candidature and improve overall OV performance. We experimentally validate the OM4OV pipeline and its cross-reference mechanism using three datasets from the Alignment Evaluation Initiative (OAEI) and exploit insights on OM used for OV tasks.
Authors: Takafumi Moriya, Takanori Ashihara, Masato Mimura, Hiroshi Sato, Kohei Matsuura, Ryo Masumura, Taichi Asami
Abstract: A hybrid autoregressive transducer (HAT) is a variant of neural transducer that models blank and non-blank posterior distributions separately. In this paper, we propose a novel internal acoustic model (IAM) training strategy to enhance HAT-based speech recognition. IAM consists of encoder and joint networks, which are fully shared and jointly trained with HAT. This joint training not only enhances the HAT training efficiency but also encourages IAM and HAT to emit blanks synchronously which skips the more expensive non-blank computation, resulting in more effective blank thresholding for faster decoding. Experiments demonstrate that the relative error reductions of the HAT with IAM compared to the vanilla HAT are statistically significant. Moreover, we introduce dual blank thresholding, which combines both HAT- and IAM-blank thresholding and a compatible decoding algorithm. This results in a 42-75% decoding speed-up with no major performance degradation.
Authors: Tengyu Xu, Eryk Helenowski, Karthik Abinav Sankararaman, Di Jin, Kaiyan Peng, Eric Han, Shaoliang Nie, Chen Zhu, Hejia Zhang, Wenxuan Zhou, Zhouhao Zeng, Yun He, Karishma Mandyam, Arya Talabzadeh, Madian Khabsa, Gabriel Cohen, Yuandong Tian, Hao Ma, Sinong Wang, Han Fang
Abstract: Reinforcement learning from human feedback (RLHF) has become the leading approach for fine-tuning large language models (LLM). However, RLHF has limitations in multi-task learning (MTL) due to challenges of reward hacking and extreme multi-objective optimization (i.e., trade-off of multiple and/or sometimes conflicting objectives). Applying RLHF for MTL currently requires careful tuning of the weights for reward model and data combinations. This is often done via human intuition and does not generalize. In this work, we introduce a novel post-training paradigm which we called Constrained Generative Policy Optimization (CGPO). The core of CGPO is Mixture of Judges (MoJ) with cost-efficient constrained policy optimization with stratification, which can identify the perfect blend in RLHF in a principled manner. It shows strong empirical results with theoretical guarantees, does not require extensive hyper-parameter tuning, and is plug-and-play in common post-training pipelines. Together, this can detect and mitigate reward hacking behaviors while reaching a pareto-optimal point across an extremely large number of objectives. Our empirical evaluations demonstrate that CGPO significantly outperforms standard RLHF algorithms like PPO and DPO across various tasks including general chat, STEM questions, instruction following, and coding. Specifically, CGPO shows improvements of 7.4% in AlpacaEval-2 (general chat), 12.5% in Arena-Hard (STEM & reasoning), and consistent gains in other domains like math and coding. Notably, PPO, while commonly used, is prone to severe reward hacking in popular coding benchmarks, which CGPO successfully addresses. This breakthrough in RLHF not only tackles reward hacking and extreme multi-objective optimization challenges but also advances the state-of-the-art in aligning general-purpose LLMs for diverse applications.
Authors: Ziyao Zhang, Yanlin Wang, Chong Wang, Jiachi Chen, Zibin Zheng
Abstract: Code generation aims to automatically generate code from input requirements, significantly enhancing development efficiency. Recent large language models (LLMs) based approaches have shown promising results and revolutionized code generation task. Despite the promising performance, LLMs often generate contents with hallucinations, especially for the code generation scenario requiring the handling of complex contextual dependencies in practical development process. Although previous study has analyzed hallucinations in LLM-powered code generation, the study is limited to standalone function generation. In this paper, we conduct an empirical study to study the phenomena, mechanism, and mitigation of LLM hallucinations within more practical and complex development contexts in repository-level generation scenario. First, we manually examine the code generation results from six mainstream LLMs to establish a hallucination taxonomy of LLM-generated code. Next, we elaborate on the phenomenon of hallucinations, analyze their distribution across different models. We then analyze causes of hallucinations and identify four potential factors contributing to hallucinations. Finally, we propose an RAG-based mitigation method, which demonstrates consistent effectiveness in all studied LLMs. The replication package including code, data, and experimental results is available at https://github.com/DeepSoftwareAnalytics/LLMCodingHallucination
URLs: https://github.com/DeepSoftwareAnalytics/LLMCodingHallucination
Authors: Haotian Zhang, Mingfei Gao, Zhe Gan, Philipp Dufter, Nina Wenzel, Forrest Huang, Dhruti Shah, Xianzhi Du, Bowen Zhang, Yanghao Li, Sam Dodge, Keen You, Zhen Yang, Aleksei Timofeev, Mingze Xu, Hong-You Chen, Jean-Philippe Fauconnier, Zhengfeng Lai, Haoxuan You, Zirui Wang, Afshin Dehghan, Peter Grasch, Yinfei Yang
Abstract: We present MM1.5, a new family of multimodal large language models (MLLMs) designed to enhance capabilities in text-rich image understanding, visual referring and grounding, and multi-image reasoning. Building upon the MM1 architecture, MM1.5 adopts a data-centric approach to model training, systematically exploring the impact of diverse data mixtures across the entire model training lifecycle. This includes high-quality OCR data and synthetic captions for continual pre-training, as well as an optimized visual instruction-tuning data mixture for supervised fine-tuning. Our models range from 1B to 30B parameters, encompassing both dense and mixture-of-experts (MoE) variants, and demonstrate that careful data curation and training strategies can yield strong performance even at small scales (1B and 3B). Additionally, we introduce two specialized variants: MM1.5-Video, designed for video understanding, and MM1.5-UI, tailored for mobile UI understanding. Through extensive empirical studies and ablations, we provide detailed insights into the training processes and decisions that inform our final designs, offering valuable guidance for future research in MLLM development.
Authors: Yukio Ohsawa
Abstract: The logical connectivity of text is represented by the connectivity of words that form archipelagoes. Here, each archipelago is a sequence of islands of the occurrences of a certain word. An island here means the local sequence of sentences where the word is emphasized, and an archipelago of a length comparable to the target text is extracted using the co-variation of entropy A (the window-based entropy) on the distribution of the word's occurrences with the width of each time window. Then, the logical connectivity of text is evaluated on entropy B (the graph-based entropy) computed on the distribution of sentences to connected word-clusters obtained on the co-occurrence of words. The results show the parts of the target text with words forming archipelagoes extracted on entropy A, without learned or prepared knowledge, form an explanatory part of the text that is of smaller entropy B than the parts extracted by the baseline methods.
Authors: Qingyu Chen, Yan Hu, Xueqing Peng, Qianqian Xie, Qiao Jin, Aidan Gilson, Maxwell B. Singer, Xuguang Ai, Po-Ting Lai, Zhizheng Wang, Vipina Kuttichi Keloth, Kalpana Raja, Jiming Huang, Huan He, Fongci Lin, Jingcheng Du, Rui Zhang, W. Jim Zheng, Ron A. Adelman, Zhiyong Lu, Hua Xu
Abstract: The biomedical literature is rapidly expanding, posing a significant challenge for manual curation and knowledge discovery. Biomedical Natural Language Processing (BioNLP) has emerged as a powerful solution, enabling the automated extraction of information and knowledge from this extensive literature. Recent attention has been directed towards Large Language Models (LLMs) due to their impressive performance. However, there remains a critical gap in understanding the effectiveness of LLMs in BioNLP tasks and their broader implications for method development and downstream users. Currently, there is a lack of baseline performance data, benchmarks, and practical recommendations for using LLMs in the biomedical domain. To address this gap, we present a systematic evaluation of four representative LLMs: GPT-3.5 and GPT-4 (closed-source), LLaMA 2 (open-sourced), and PMC LLaMA (domain-specific) across 12 BioNLP datasets covering six applications (named entity recognition, relation extraction, multi-label document classification, question answering, text summarization, and text simplification). The evaluation is conducted under four settings: zero-shot, static few-shot, dynamic K-nearest few-shot, and fine-tuning. We compare these models against state-of-the-art (SOTA) approaches that fine-tune (domain-specific) BERT or BART models, which are well-established methods in BioNLP tasks. The evaluation covers both quantitative and qualitative evaluations, where the latter involves manually reviewing collectively hundreds of thousands of LLM outputs for inconsistencies, missing information, and hallucinations in extractive and classification tasks. The qualitative review also examines accuracy, 1 completeness, and readability in text summarization tasks. Additionally, a cost analysis of closed-source GPT models is conducted.
Authors: Jacob-Junqi Tian, Omkar Dige, D. B. Emerson, Faiza Khan Khattak
Abstract: Large language models (LLMs) are trained on vast, uncurated datasets that contain various forms of biases and language reinforcing harmful stereotypes that may be subsequently inherited by the models themselves. Therefore, it is essential to examine and address biases in language models, integrating fairness into their development to ensure that these models do not perpetuate social biases. In this work, we demonstrate the importance of reasoning in zero-shot stereotype identification across several open-source LLMs. Accurate identification of stereotypical language is a complex task requiring a nuanced understanding of social structures, biases, and existing unfair generalizations about particular groups. While improved accuracy is observed through model scaling, the use of reasoning, especially multi-step reasoning, is crucial to consistent performance. Additionally, through a qualitative analysis of select reasoning traces, we highlight how reasoning improves not just accuracy, but also the interpretability of model decisions. This work firmly establishes reasoning as a critical component in automatic stereotype detection and is a first step towards stronger stereotype mitigation pipelines for LLMs.
Authors: Hong Zhang, Quoc-Nam Nguyen, Prasanta Bhattacharya, Wei Gao, Liang Ze Wong, Brandon Siyuan Loh, Joseph J. P. Simons, Jisun An
Abstract: This study enhances stance detection on social media by incorporating deeper psychological attributes, specifically individuals' moral foundations. These theoretically-derived dimensions aim to provide a comprehensive profile of an individual's moral concerns which, in recent work, has been linked to behaviour in a range of domains, including society, politics, health, and the environment. In this paper, we investigate how moral foundation dimensions can contribute to predicting an individual's stance on a given target. Specifically we incorporate moral foundation features extracted from text, along with message semantic features, to classify stances at both message- and user-levels using both traditional machine learning models and large language models. Our preliminary results suggest that encoding moral foundations can enhance the performance of stance detection tasks and help illuminate the associations between specific moral foundations and online stances on target topics. The results highlight the importance of considering deeper psychological attributes in stance analysis and underscores the role of moral foundations in guiding online social behavior.
Authors: Haoran Li, Yulin Chen, Jinglong Luo, Jiecong Wang, Hao Peng, Yan Kang, Xiaojin Zhang, Qi Hu, Chunkit Chan, Zenglin Xu, Bryan Hooi, Yangqiu Song
Abstract: The advancement of large language models (LLMs) has significantly enhanced the ability to effectively tackle various downstream NLP tasks and unify these tasks into generative pipelines. On the one hand, powerful language models, trained on massive textual data, have brought unparalleled accessibility and usability for both models and users. On the other hand, unrestricted access to these models can also introduce potential malicious and unintentional privacy risks. Despite ongoing efforts to address the safety and privacy concerns associated with LLMs, the problem remains unresolved. In this paper, we provide a comprehensive analysis of the current privacy attacks targeting LLMs and categorize them according to the adversary's assumed capabilities to shed light on the potential vulnerabilities present in LLMs. Then, we present a detailed overview of prominent defense strategies that have been developed to counter these privacy attacks. Beyond existing works, we identify upcoming privacy concerns as LLMs evolve. Lastly, we point out several potential avenues for future exploration.
Authors: Chia-Hsuan Lee, Hao Cheng, Mari Ostendorf
Abstract: Large language models (LLMs) have revolutionized the landscape of Natural Language Processing systems, but are computationally expensive. To reduce the cost without sacrificing performance, previous studies have explored various approaches to harness the potential of Small Language Models (SLMs) as cost-effective alternatives to their larger counterparts. Driven by findings that SLMs and LLMs exhibit complementary strengths in a structured knowledge extraction task, this work presents a novel SLM/LLM routing framework designed to improve computational efficiency and enhance task performance. First, exemplar pools are created to represent the types of contexts where each LM provides a more reliable answer, leveraging a sentence embedding fine-tuned so that context similarity is close to dialogue state similarity. Then, during inference, the k-nearest exemplars to the testing instance are retrieved, and the instance is routed according to majority vote. In dialogue state tracking tasks, the proposed routing framework enhances performance substantially compared to relying solely on LLMs, while reducing the computational costs by over 50%.
Authors: Prakash Chandra Sukhwal, Vaibhav Rajan, Atreyi Kankanhalli
Abstract: Medical question answer (QA) assistants respond to lay users' health-related queries by synthesizing information from multiple sources using natural language processing and related techniques. They can serve as vital tools to alleviate issues of misinformation, information overload, and complexity of medical language, thus addressing lay users' information needs while reducing the burden on healthcare professionals. QA systems, the engines of such assistants, have typically used either language models (LMs) or knowledge graphs (KG), though the approaches could be complementary. LM-based QA systems excel at understanding complex questions and providing well-formed answers, but are prone to factual mistakes. KG-based QA systems, which represent facts well, are mostly limited to answering short-answer questions with pre-created templates. While a few studies have jointly used LM and KG approaches for text-based QA, this was done to answer multiple-choice questions. Extant QA systems also have limitations in terms of automation and performance. We address these challenges by designing a novel, automated disease QA system which effectively utilizes both LM and KG techniques through a joint-reasoning approach to answer disease-related questions appropriate for lay users. Our evaluation of the system using a range of quality metrics demonstrates its efficacy over benchmark systems, including the popular ChatGPT.
Authors: Isabel Cachola, Silviu Cucerzan, Allen Herring, Vuksan Mijovic, Erik Oveson, Sujay Kumar Jauhar
Abstract: Authors seeking to communicate with broader audiences often share their ideas in various document formats, such as slide decks, newsletters, reports, and posters. Prior work on document generation has generally tackled the creation of each separate format to be a different task, leading to fragmented learning processes, redundancy in models and methods, and disjointed evaluation. We consider each of these documents as templatic views of the same underlying knowledge/content, and we aim to unify the generation and evaluation of these templatic views. We begin by showing that current LLMs are capable of generating various document formats with little to no supervision. Further, a simple augmentation involving a structured intermediate representation can improve performance, especially for smaller models. We then introduce a novel unified evaluation framework that can be adapted to measuring the quality of document generators for heterogeneous downstream applications. This evaluation is adaptable to a range of user defined criteria and application scenarios, obviating the need for task specific evaluation metrics. Finally, we conduct a human evaluation, which shows that people prefer 82% of the documents generated with our method, while correlating more highly with our unified evaluation framework than prior metrics in the literature.
Authors: Haritz Puerto, Martin Tutek, Somak Aditya, Xiaodan Zhu, Iryna Gurevych
Abstract: Reasoning is a fundamental component of language understanding. Recent prompting techniques, such as chain of thought, have consistently improved LLMs' performance on various reasoning tasks. Nevertheless, there is still little understanding of what triggers reasoning abilities in LLMs in the inference stage. In this paper, we introduce code prompting, a chain of prompts that transforms a natural language problem into code and directly prompts the LLM using the generated code without resorting to external code execution. We hypothesize that code prompts can elicit certain reasoning capabilities of LLMs trained on text and code and utilize the proposed method to improve conditional reasoning, the ability to infer different conclusions depending on the fulfillment of certain conditions. We find that code prompting exhibits a high-performance boost for multiple LLMs (up to 22.52 percentage points on GPT 3.5, 7.75 on Mixtral, and 16.78 on Mistral) across multiple conditional reasoning datasets. We then conduct comprehensive experiments to understand how code prompts trigger reasoning abilities and which capabilities are elicited in the underlying models. Our analysis of GPT 3.5 reveals that the code formatting of the input problem is essential for performance improvement. Furthermore, code prompts improve sample efficiency of in-context learning and facilitate state tracking of variables or entities.
Authors: Yuan He, Zhangdie Yuan, Jiaoyan Chen, Ian Horrocks
Abstract: Interpreting hierarchical structures latent in language is a key limitation of current language models (LMs). While previous research has implicitly leveraged these hierarchies to enhance LMs, approaches for their explicit encoding are yet to be explored. To address this, we introduce a novel approach to re-train transformer encoder-based LMs as Hierarchy Transformer encoders (HiTs), harnessing the expansive nature of hyperbolic space. Our method situates the output embedding space of pre-trained LMs within a Poincar\'e ball with a curvature that adapts to the embedding dimension, followed by training on hyperbolic clustering and centripetal losses. These losses are designed to effectively cluster related entities (input as texts) and organise them hierarchically. We evaluate HiTs against pre-trained LMs, standard fine-tuned LMs, and several hyperbolic embedding baselines, focusing on their capabilities in simulating transitive inference, predicting subsumptions, and transferring knowledge across hierarchies. The results demonstrate that HiTs consistently outperform all baselines in these tasks, underscoring the effectiveness and transferability of our re-trained hierarchy encoders.
Authors: Fengbin Zhu, Ziyang Liu, Fuli Feng, Chao Wang, Moxin Li, Tat-Seng Chua
Abstract: In this work, we address question answering (QA) over a hybrid of tabular and textual data that are very common content on the Web (e.g. SEC filings), where discrete reasoning capabilities are often required. Recently, large language models (LLMs) like GPT-4 have demonstrated strong multi-step reasoning capabilities. We then consider harnessing the amazing power of LLMs to solve our task. We abstract a Step-wise Pipeline for tabular and textual QA, which consists of three key steps, including Extractor, Reasoner and Executor, and initially design an instruction to instantiate the pipeline and validate that GPT-4 outperforms all existing methods. However, utilizing an online LLM like GPT-4 holds various challenges in terms of cost, latency, and data security risk, which motivates us to specialize smaller LLMs in this task. We develop a TAT-LLM language model by fine-tuning LLaMA 2 with the training data generated automatically from existing expert-annotated datasets following the Step-wise Pipeline. The experimental results have verified that our TAT-LLM model can outperform all baseline models, including the previous best fine-tuned models and very large-scale LLMs like GPT-4 on FinQA, TAT-QA and TAT-DQA benchmarks.
Authors: Amir Taubenfeld, Yaniv Dover, Roi Reichart, Ariel Goldstein
Abstract: The emergence of Large Language Models (LLMs), has opened exciting possibilities for constructing computational simulations designed to replicate human behavior accurately. Current research suggests that LLM-based agents become increasingly human-like in their performance, sparking interest in using these AI agents as substitutes for human participants in behavioral studies. However, LLMs are complex statistical learners without straightforward deductive rules, making them prone to unexpected behaviors. Hence, it is crucial to study and pinpoint the key behavioral distinctions between humans and LLM-based agents. In this study, we highlight the limitations of LLMs in simulating human interactions, particularly focusing on LLMs' ability to simulate political debates on topics that are important aspects of people's day-to-day lives and decision-making processes. Our findings indicate a tendency for LLM agents to conform to the model's inherent social biases despite being directed to debate from certain political perspectives. This tendency results in behavioral patterns that seem to deviate from well-established social dynamics among humans. We reinforce these observations using an automatic self-fine-tuning method, which enables us to manipulate the biases within the LLM and demonstrate that agents subsequently align with the altered biases. These results underscore the need for further research to develop methods that help agents overcome these biases, a critical step toward creating more realistic simulations.
Authors: Hanxing Ding, Liang Pang, Zihao Wei, Huawei Shen, Xueqi Cheng
Abstract: Hallucinations pose a significant challenge for the practical implementation of large language models (LLMs). The utilization of parametric knowledge in generating factual content is constrained by the limited knowledge of LLMs, potentially resulting in internal hallucinations. While incorporating external information can help fill knowledge gaps, it also introduces the risk of irrelevant information, thereby increasing the likelihood of external hallucinations. A careful and balanced integration of the parametric knowledge within LLMs with external information is crucial to alleviate hallucinations. In this study, we present Rowen, a novel approach that enhances LLMs with a selective retrieval augmentation process tailored to address hallucinated outputs. This process is governed by a multilingual semantic-aware detection module, which evaluates the consistency of the perturbed responses across various languages for the same queries. Upon detecting inconsistencies indicative of hallucinations, Rowen activates the retrieval of external information to rectify the model outputs. Rowen adeptly harmonizes the intrinsic parameters in LLMs with external knowledge sources, effectively mitigating hallucinations by ensuring a balanced integration of internal reasoning and external evidence. Through a comprehensive empirical analysis, we demonstrate that Rowen surpasses the current state-of-the-art in both detecting and mitigating hallucinated content within the outputs of LLMs.
Authors: Zongxia Li, Ishani Mondal, Yijun Liang, Huy Nghiem, Jordan Lee Boyd-Graber
Abstract: Question answering (QA) can only make progress if we know if an answer is correct, but current answer correctness (AC) metrics struggle with verbose, free-form answers from large language models (LLMs). There are two challenges with current short-form QA evaluations: a lack of diverse styles of evaluation data and an over-reliance on expensive and slow LLMs. LLM-based scorers correlate better with humans, but this expensive task has only been tested on limited QA datasets. We rectify these issues by providing rubrics and datasets for evaluating machine QA adopted from the Trivia community. We also propose an efficient, and interpretable QA evaluation that is more stable than an exact match and neural methods(BERTScore).
Authors: Mingxu Tao, Dongyan Zhao, Yansong Feng
Abstract: Open-ended question answering requires models to find appropriate evidence to form well-reasoned, comprehensive and helpful answers. In practical applications, models also need to engage in extended discussions on potential scenarios closely relevant to the question. With augmentation of retrieval module, open-source Large Language Models (LLMs) can produce coherent answers often with different focuses, but are still sub-optimal in terms of reliable evidence selection and in-depth question analysis. In this paper, we propose a novel Chain-of-Discussion framework to leverage the synergy among multiple open-source LLMs aiming to provide \textbf{more correct} and \textbf{more comprehensive} answers for open-ended QA, although they are not strong enough individually. Our experiments show that discussions among multiple LLMs play a vital role in enhancing the quality of answers. We release our data and code at \url{https://github.com/kobayashikanna01/Chain-of-Discussion}.
URLs: https://github.com/kobayashikanna01/Chain-of-Discussion
Authors: Ayana Niwa, Hayate Iso
Abstract: We introduce AmbigNLG, a novel task designed to tackle the challenge of task ambiguity in instructions for Natural Language Generation (NLG). Ambiguous instructions often impede the performance of Large Language Models (LLMs), especially in complex NLG tasks. To tackle this issue, we propose an ambiguity taxonomy that categorizes different types of instruction ambiguities and refines initial instructions with clearer specifications. Accompanying this task, we present AmbigSNI-NLG, a dataset comprising 2,500 instances annotated to facilitate research in AmbigNLG. Through comprehensive experiments with state-of-the-art LLMs, we demonstrate that our method significantly enhances the alignment of generated text with user expectations, achieving up to a 15.02-point increase in ROUGE scores. Our findings highlight the critical importance of addressing task ambiguity to fully harness the capabilities of LLMs in NLG tasks. Furthermore, we confirm the effectiveness of our method in practical settings involving interactive ambiguity mitigation with users, underscoring the benefits of leveraging LLMs for interactive clarification.
Authors: Swagata Ashwani, Kshiteesh Hegde, Nishith Reddy Mannuru, Mayank Jindal, Dushyant Singh Sengar, Krishna Chaitanya Rao Kathala, Dishant Banga, Vinija Jain, Aman Chadha
Abstract: With the rise of Large Language Models(LLMs), it has become crucial to understand their capabilities and limitations in deciphering and explaining the complex web of causal relationships that language entails. Current methods use either explicit or implicit causal reasoning, yet there is a strong need for a unified approach combining both to tackle a wide array of causal relationships more effectively. This research proposes a novel architecture called Context Aware Reasoning Enhancement with Counterfactual Analysis(CARE CA) framework to enhance causal reasoning and explainability. The proposed framework incorporates an explicit causal detection module with ConceptNet and counterfactual statements, as well as implicit causal detection through LLMs. Our framework goes one step further with a layer of counterfactual explanations to accentuate LLMs understanding of causality. The knowledge from ConceptNet enhances the performance of multiple causal reasoning tasks such as causal discovery, causal identification and counterfactual reasoning. The counterfactual sentences add explicit knowledge of the not caused by scenarios. By combining these powerful modules, our model aims to provide a deeper understanding of causal relationships, enabling enhanced interpretability. Evaluation of benchmark datasets shows improved performance across all metrics, such as accuracy, precision, recall, and F1 scores. We also introduce CausalNet, a new dataset accompanied by our code, to facilitate further research in this domain.
Authors: Shitong Duan, Xiaoyuan Yi, Peng Zhang, Yan Liu, Zheng Liu, Tun Lu, Xing Xie, Ning Gu
Abstract: Large language models (LLMs) have revolutionized the role of AI, yet pose potential social risks. To steer LLMs towards human preference, alignment technologies have been introduced and gained increasing attention. Nevertheless, existing methods heavily rely on high-quality positive-negative training pairs, suffering from noisy positive responses that are barely distinguishable from negative ones. Given recent LLMs' proficiency in generating helpful responses, this work pivots towards a new research question: can we achieve alignment using solely human-annotated negative samples, preserving helpfulness while reducing harmfulness? For this purpose, we propose Distributional Dispreference Optimization (D$^2$O), which maximizes the discrepancy between dispreferred responses and the generated non-negative ones. In this way, D$^2$O effectively eschews harmful information without incorporating noisy positive samples, while avoiding collapse using self-generated responses as anchors. We demonstrate that D$^2$O can be regarded as learning a distributional preference model reflecting human dispreference against negative responses, which is theoretically an upper bound of the instance-level DPO. Extensive experiments manifest that our method achieves comparable generation quality and surpasses the latest strong baselines in producing less harmful and more informative responses with better training stability and faster convergence.
Authors: Yekyung Kim, Yapei Chang, Marzena Karpinska, Aparna Garimella, Varun Manjunatha, Kyle Lo, Tanya Goyal, Mohit Iyyer
Abstract: While long-context large language models (LLMs) can technically summarize book-length documents (>100K tokens), the length and complexity of the documents have so far prohibited evaluations of input-dependent aspects like faithfulness. In this paper, we conduct the first large-scale human evaluation of faithfulness and content selection on LLM-generated summaries of fictional books. Our study mitigates the issue of data contamination by focusing on summaries of books published in 2023 or 2024, and we hire annotators who have fully read each book prior to the annotation task to minimize cost and cognitive burden. We collect FABLES, a dataset of annotations on 3,158 claims made in LLM-generated summaries of 26 books, at a cost of $5.2K USD, which allows us to rank LLM summarizers based on faithfulness: Claude-3-Opus significantly outperforms all closed-source LLMs, while the open-source Mixtral is on par with GPT-3.5-Turbo. An analysis of the annotations reveals that most unfaithful claims relate to events and character states, and they generally require indirect reasoning over the narrative to invalidate. While LLM-based auto-raters have proven reliable for factuality and coherence in other settings, we implement several LLM raters of faithfulness and find that none correlates strongly with human annotations, especially with regard to detecting unfaithful claims. Our experiments suggest that detecting unfaithful claims is an important future direction not only for summarization evaluation but also as a testbed for long-context understanding. Finally, we move beyond faithfulness by exploring content selection errors in book-length summarization: we develop a typology of omission errors related to crucial narrative elements and also identify a systematic over-emphasis on events occurring towards the end of the book.
Authors: Yanpeng Ye, Jie Ren, Shaozhou Wang, Yuwei Wan, Haofen Wang, Imran Razzak, Bram Hoex, Tong Xie, Wenjie Zhang
Abstract: Knowledge in materials science is widely dispersed across extensive scientific literature, posing significant challenges for efficient discovery and integration of new materials. Traditional methods, often reliant on costly and time-consuming experimental approaches, further complicate rapid innovation. Addressing these challenges, the integration of artificial intelligence with materials science has opened avenues for accelerating the discovery process, though it also demands precise annotation, data extraction, and traceability of information. To tackle these issues, this article introduces the Materials Knowledge Graph (MKG), which utilizes advanced natural language processing techniques, integrated with large language models to extract and systematically organize a decade's worth of high-quality research into structured triples, contains 162,605 nodes and 731,772 edges. MKG categorizes information into comprehensive labels such as Name, Formula, and Application, structured around a meticulously designed ontology, thus enhancing data usability and integration. By implementing network-based algorithms, MKG not only facilitates efficient link prediction but also significantly reduces reliance on traditional experimental methods. This structured approach not only streamlines materials research but also lays the groundwork for more sophisticated science knowledge graphs.
Authors: Xianlong Zeng, Yijing Gao, Fanghao Song, Ang Liu
Abstract: This study introduces a simple yet effective method for identifying similar data points across non-free text domains, such as tabular and image data, using Large Language Models (LLMs). Our two-step approach involves data point summarization and hidden state extraction. Initially, data is condensed via summarization using an LLM, reducing complexity and highlighting essential information in sentences. Subsequently, the summarization sentences are fed through another LLM to extract hidden states, serving as compact, feature-rich representations. This approach leverages the advanced comprehension and generative capabilities of LLMs, offering a scalable and efficient strategy for similarity identification across diverse datasets. We demonstrate the effectiveness of our method in identifying similar data points on multiple datasets. Additionally, our approach enables non-technical domain experts, such as fraud investigators or marketing operators, to quickly identify similar data points tailored to specific scenarios, demonstrating its utility in practical applications. In general, our results open new avenues for leveraging LLMs in data analysis across various domains
Authors: Jiao Ou, Jiayu Wu, Che Liu, Fuzheng Zhang, Di Zhang, Kun Gai
Abstract: Aligning large language models (LLMs) with human expectations requires high-quality instructional dialogues, which usually require instructions that are diverse and in-depth. Existing methods leverage two LLMs to interact for automatic collection: one simulating a user to pose instructions, and the other acting as a system agent to respond. However, these user simulators struggle to model the rules behind how dialogues can pose different instructions without explicit guidance, resulting in general instructions. In this paper, we propose to explicitly capture the complex rules to help the user simulator pose diverse and in-depth instruction. Specifically, we first induce high-level instruction strategies from various real instruction dialogues serving as rules. Afterward, different possible strategies are applied to the newly given dialogue scenario deductively to pose various instructions. Experimental results show that our method can generate diverse and in-depth instructions. The constructed multi-turn instructional dialogues can outperform competitive baselines on the downstream chat model.
Authors: Shuchang Tao, Liuyi Yao, Hanxing Ding, Yuexiang Xie, Qi Cao, Fei Sun, Jinyang Gao, Huawei Shen, Bolin Ding
Abstract: Despite the success of large language models (LLMs) in natural language generation, much evidence shows that LLMs may produce incorrect or nonsensical text. This limitation highlights the importance of discerning when to trust LLMs, especially in safety-critical domains. Existing methods often express reliability by confidence level, however, their effectiveness is limited by the lack of objective guidance. To address this, we propose CONfidence-Quality-ORDer-preserving alignment approach (CONQORD), which leverages reinforcement learning guided by a tailored dual-component reward function. This function integrates quality reward and order-preserving alignment reward functions. Specifically, the order-preserving reward incentivizes the model to verbalize greater confidence for responses of higher quality to align the order of confidence and quality. Experiments demonstrate that CONQORD significantly improves the alignment performance between confidence and response accuracy, without causing over-cautious. Furthermore, the aligned confidence provided by CONQORD informs when to trust LLMs, and acts as a determinant for initiating the retrieval process of external knowledge. Aligning confidence with response quality ensures more transparent and reliable responses, providing better trustworthiness.
Authors: Xiaodong Liu
Abstract: This paper presents a significant improvement on the previous conference paper known as DefSent. The prior study seeks to improve sentence embeddings of language models by projecting definition sentences into the vector space of dictionary entries. We discover that this approach is not fully explored due to the methodological limitation of using word embeddings of language models to represent dictionary entries. This leads to two hindrances. First, dictionary entries are constrained by the single-word vocabulary, and thus cannot be fully exploited. Second, semantic representations of language models are known to be anisotropic, but pre-processing word embeddings for DefSent is not allowed because its weight is frozen during training and tied to the prediction layer. In this paper, we propose a novel method to progressively build entry embeddings not subject to the limitations. As a result, definition sentences can be projected into a quasi-isotropic or isotropic vector space of unlimited dictionary entries, so that sentence embeddings of noticeably better quality are attainable. We abbreviate our approach as DefSent+ (a plus version of DefSent), involving the following strengths: 1) the task performance on measuring sentence similarities is significantly improved compared to DefSent; 2) when DefSent+ is used to further train data-augmented models like SIMCSE, SNCSE, and SynCSE, state-of-the-art performance on measuring sentence similarities can be achieved among the approaches without using manually labeled datasets; 3) DefSent+ is also competitive in feature-based transfer for NLP downstream tasks.
Authors: Chenyang Le, Yao Qian, Dongmei Wang, Long Zhou, Shujie Liu, Xiaofei Wang, Midia Yousefi, Yanmin Qian, Jinyu Li, Sheng Zhao, Michael Zeng
Abstract: There is a rising interest and trend in research towards directly translating speech from one language to another, known as end-to-end speech-to-speech translation. However, most end-to-end models struggle to outperform cascade models, i.e., a pipeline framework by concatenating speech recognition, machine translation and text-to-speech models. The primary challenges stem from the inherent complexities involved in direct translation tasks and the scarcity of data. In this study, we introduce a novel model framework TransVIP that leverages diverse datasets in a cascade fashion yet facilitates end-to-end inference through joint probability. Furthermore, we propose two separated encoders to preserve the speaker's voice characteristics and isochrony from the source speech during the translation process, making it highly suitable for scenarios such as video dubbing. Our experiments on the French-English language pair demonstrate that our model outperforms the current state-of-the-art speech-to-speech translation model.
Authors: Linxin Song, Jiale Liu, Jieyu Zhang, Shaokun Zhang, Ao Luo, Shijian Wang, Qingyun Wu, Chi Wang
Abstract: Leveraging multiple large language model (LLM) agents has shown to be a promising approach for tackling complex tasks, while the effective design of multiple agents for a particular application remains an art. It is thus intriguing to answer a critical question: Given a task, how can we build a team of LLM agents to solve it effectively? Our new adaptive team-building paradigm offers a flexible solution, realized through a novel agent design named Captain Agent. It dynamically forms and manages teams for each step of a task-solving process, utilizing nested group conversations and reflection to ensure diverse expertise and prevent stereotypical outputs, allowing for a flexible yet structured approach to problem-solving. A comprehensive evaluation across six real-world scenarios demonstrates that Captain Agent significantly outperforms existing multi-agent methods with 21.94% improvement in average accuracy, providing outstanding performance without requiring task-specific prompt engineering. Our exploration of different backbone LLM and cost analysis further shows that Captain Agent can improve the conversation quality of weak LLM and achieve competitive performance with extremely low cost, which illuminates the application of multi-agent systems.
Authors: Chen Zhang, Chengguang Tang, Dading Chong, Ke Shi, Guohua Tang, Feng Jiang, Haizhou Li
Abstract: Mainstream approaches to aligning large language models (LLMs) heavily rely on human preference data, particularly when models require periodic updates. The standard process for iterative alignment of LLMs involves collecting new human feedback for each update. However, the data collection process is costly and challenging to scale. To address this issue, we introduce the "TS-Align" framework, which fine-tunes a policy model using pairwise feedback data automatically mined from its outputs. This automatic mining process is efficiently accomplished through the collaboration between a large-scale teacher model and a small-scale student model. The policy fine-tuning process can be iteratively repeated using on-policy generations within our proposed teacher-student collaborative framework. Through extensive experiments, we demonstrate that our final aligned policy outperforms the base policy model with an average win rate of 69.7% across seven conversational or instruction-following datasets. Furthermore, we show that the ranking capability of the teacher is effectively distilled into the student through our pipeline, resulting in a small-scale yet effective reward model for policy model alignment.
Authors: Dojun Park, Jiwoo Lee, Seohyun Park, Hyeyun Jeong, Youngeun Koo, Soonha Hwang, Seonwoo Park, Sungeun Lee
Abstract: As the capabilities of Large Language Models (LLMs) expand, it becomes increasingly important to evaluate them beyond basic knowledge assessment, focusing on higher-level language understanding. This study introduces MultiPragEval, the first multilingual pragmatic evaluation of LLMs, designed for English, German, Korean, and Chinese. Comprising 1200 question units categorized according to Grice's Cooperative Principle and its four conversational maxims, MultiPragEval enables an in-depth assessment of LLMs' contextual awareness and their ability to infer implied meanings. Our findings demonstrate that Claude3-Opus significantly outperforms other models in all tested languages, establishing a state-of-the-art in the field. Among open-source models, Solar-10.7B and Qwen1.5-14B emerge as strong competitors. By analyzing pragmatic inference, we provide valuable insights into the capabilities essential for advanced language comprehension in AI systems.
Authors: Lin Shi, Chiyu Ma, Wenhua Liang, Weicheng Ma, Soroush Vosoughi
Abstract: LLM-as-a-Judge presents a promising alternative to human evaluators across various tasks, but inherent biases, especially position bias - a tendency to favor solutions based on their position in the prompt - have compromised its effectiveness. Our study introduces a systematic framework to examine position bias in pairwise comparisons, focusing on repetition stability, position consistency, and preference fairness. This research significantly contributes to the field by introducing new concepts for understanding position bias and providing a multi-dimensional framework for evaluations. We conducted experiments with 12 LLM judges across MTBench and DevBench, covering 22 tasks and approximately 40 solution-generating models - candidates, resulting in over 100,000 evaluation instances. Our findings confirm that position bias in capable LLM judges is not due to random chances, along with notable variations observed across judges and tasks. Moreover, position bias is weakly influenced by the length of prompt components but significantly impacted by the quality gap between solutions. These insights can help optimize judge model selections, improve benchmark design, and inform future research on debiasing strategies, ultimately enhancing the reliability of LLM judges.
Authors: Somin Wadhwa, Adit Krishnan, Runhui Wang, Byron C. Wallace, Chris Kong
Abstract: Entity matching is the task of linking records from different sources that refer to the same real-world entity. Past work has primarily treated entity linking as a standard supervised learning problem. However, supervised entity matching models often do not generalize well to new data, and collecting exhaustive labeled training data is often cost prohibitive. Further, recent efforts have adopted LLMs for this task in few/zero-shot settings, exploiting their general knowledge. But LLMs are prohibitively expensive for performing inference at scale for real-world entity matching tasks. As an efficient alternative, we re-cast entity matching as a conditional generation task as opposed to binary classification. This enables us to "distill" LLM reasoning into smaller entity matching models via natural language explanations. This approach achieves strong performance, especially on out-of-domain generalization tests (10.85% F-1) where standalone generative methods struggle. We perform ablations that highlight the importance of explanations, both for performance and model robustness.
Authors: Wenxuan Ding, Weiqi Wang, Sze Heng Douglas Kwok, Minghao Liu, Tianqing Fang, Jiaxin Bai, Xin Liu, Changlong Yu, Zheng Li, Chen Luo, Qingyu Yin, Bing Yin, Junxian He, Yangqiu Song
Abstract: Enhancing Language Models' (LMs) ability to understand purchase intentions in E-commerce scenarios is crucial for their effective assistance in various downstream tasks. However, previous approaches that distill intentions from LMs often fail to generate meaningful and human-centric intentions applicable in real-world E-commerce contexts. This raises concerns about the true comprehension and utilization of purchase intentions by LMs. In this paper, we present IntentionQA, a double-task multiple-choice question answering benchmark to evaluate LMs' comprehension of purchase intentions in E-commerce. Specifically, LMs are tasked to infer intentions based on purchased products and utilize them to predict additional purchases. IntentionQA consists of 4,360 carefully curated problems across three difficulty levels, constructed using an automated pipeline to ensure scalability on large E-commerce platforms. Human evaluations demonstrate the high quality and low false-negative rate of our benchmark. Extensive experiments across 19 language models show that they still struggle with certain scenarios, such as understanding products and intentions accurately, jointly reasoning with products and intentions, and more, in which they fall far behind human performances. Our code and data are publicly available at https://github.com/HKUST-KnowComp/IntentionQA.
Authors: Yu Zhang, Xiusi Chen, Bowen Jin, Sheng Wang, Shuiwang Ji, Wei Wang, Jiawei Han
Abstract: In many scientific fields, large language models (LLMs) have revolutionized the way text and other modalities of data (e.g., molecules and proteins) are handled, achieving superior performance in various applications and augmenting the scientific discovery process. Nevertheless, previous surveys on scientific LLMs often concentrate on one or two fields or a single modality. In this paper, we aim to provide a more holistic view of the research landscape by unveiling cross-field and cross-modal connections between scientific LLMs regarding their architectures and pre-training techniques. To this end, we comprehensively survey over 260 scientific LLMs, discuss their commonalities and differences, as well as summarize pre-training datasets and evaluation tasks for each field and modality. Moreover, we investigate how LLMs have been deployed to benefit scientific discovery. Resources related to this survey are available at https://github.com/yuzhimanhua/Awesome-Scientific-Language-Models.
URLs: https://github.com/yuzhimanhua/Awesome-Scientific-Language-Models.
Authors: Byung-Doh Oh, William Schuler
Abstract: Predictions of word-by-word conditional probabilities from Transformer-based language models are often evaluated to model the incremental processing difficulty of human readers. In this paper, we argue that there is a confound posed by the most common method of aggregating subword probabilities of such language models into word probabilities. This is due to the fact that tokens in the subword vocabulary of most language models have leading whitespaces and therefore do not naturally define stop probabilities of words. We first prove that this can result in distributions over word probabilities that sum to more than one, thereby violating the axiom that $\mathsf{P}(\Omega) = 1$. This property results in a misallocation of word-by-word surprisal, where the unacceptability of the end of the current word is incorrectly carried over to the next word. Additionally, this implicit prediction of word boundaries incorrectly models psycholinguistic experiments where human subjects directly observe upcoming word boundaries. We present a simple decoding technique to reaccount the probability of the trailing whitespace into that of the current word, which resolves this confound. Experiments show that this correction reveals lower estimates of garden-path effects in transitive/intransitive sentences and poorer fits to naturalistic reading times.
Authors: Junru Lu, Jiazheng Li, Siyu An, Meng Zhao, Yulan He, Di Yin, Xing Sun
Abstract: Direct Preference Optimization (DPO) has emerged as a prominent algorithm for the direct and robust alignment of Large Language Models (LLMs) with human preferences, offering a more straightforward alternative to the complex Reinforcement Learning from Human Feedback (RLHF). Despite its promising efficacy, DPO faces a notable drawback: "verbosity", a common over-optimization phenomenon also observed in RLHF. While previous studies mainly attributed verbosity to biased labels within the data, we propose that the issue also stems from an inherent algorithmic length reliance in DPO. Specifically, we suggest that the discrepancy between sequence-level Kullback-Leibler (KL) divergences between chosen and rejected sequences, used in DPO, results in overestimated or underestimated rewards due to varying token lengths. Empirically, we utilize datasets with different label lengths to demonstrate the presence of biased rewards. We then introduce an effective downsampling approach, named SamPO, to eliminate potential length reliance. Our experimental evaluations, conducted across three LLMs of varying scales and a diverse array of conditional and open-ended benchmarks, highlight the efficacy of SamPO in mitigating verbosity, achieving improvements of 5% to 12% over DPO through debaised rewards. Our codes can be accessed at: https://github.com/LuJunru/SamPO/.
Authors: Wenyan Li, Xinyu Zhang, Jiaang Li, Qiwei Peng, Raphael Tang, Li Zhou, Weijia Zhang, Guimin Hu, Yifei Yuan, Anders S{\o}gaard, Daniel Hershcovich, Desmond Elliott
Abstract: Food is a rich and varied dimension of cultural heritage, crucial to both individuals and social groups. To bridge the gap in the literature on the often-overlooked regional diversity in this domain, we introduce FoodieQA, a manually curated, fine-grained image-text dataset capturing the intricate features of food cultures across various regions in China. We evaluate vision-language Models (VLMs) and large language models (LLMs) on newly collected, unseen food images and corresponding questions. FoodieQA comprises three multiple-choice question-answering tasks where models need to answer questions based on multiple images, a single image, and text-only descriptions, respectively. While LLMs excel at text-based question answering, surpassing human accuracy, the open-sourced VLMs still fall short by 41% on multi-image and 21% on single-image VQA tasks, although closed-weights models perform closer to human levels (within 10%). Our findings highlight that understanding food and its cultural implications remains a challenging and under-explored direction.
Authors: Guan-Ting Lin, Hung-yi Lee
Abstract: Emphasis is a crucial component in human communication, which indicates the speaker's intention and implication beyond pure text in dialogue. While Large Language Models (LLMs) have revolutionized natural language processing, their ability to understand emphasis in dialogue remains unclear. This paper introduces Emphasized-Talk, a benchmark with emphasis-annotated dialogue samples capturing the implications of emphasis. We evaluate various LLMs, both open-source and commercial, to measure their performance in understanding emphasis. Additionally, we propose an automatic evaluation pipeline using GPT-4, which achieves a high correlation with human rating. Our findings reveal that although commercial LLMs generally perform better, there is still significant room for improvement in comprehending emphasized sentences.
Authors: Wanli Yang, Fei Sun, Jiajun Tan, Xinyu Ma, Du Su, Dawei Yin, Huawei Shen
Abstract: Despite significant progress in model editing methods, their application in real-world scenarios remains challenging as they often cause large language models (LLMs) to collapse. Among them, ROME is particularly concerning, as it could disrupt LLMs with only a single edit. In this paper, we study the root causes of such collapse. Through extensive analysis, we identify two primary factors that contribute to the collapse: i) inconsistent handling of prefixed and unprefixed keys in the parameter update equation may result in very small denominators, causing excessively large parameter updates; ii) the subject of collapse cases is usually the first token, whose unprefixed key distribution significantly differs from the prefixed key distribution in autoregressive transformers, causing the aforementioned issue to materialize. To validate our findings, we propose a simple yet effective approach: uniformly using prefixed keys during editing phase and adding prefixes during testing phase to ensure the consistency between training and testing. The experimental results show that the proposed solution can prevent model collapse while maintaining the effectiveness of the edits.
Authors: Kyle Moore, Jesse Roberts, Thao Pham, Oseremhen Ewaleifoh, Doug Fisher
Abstract: Cloze testing is a common method for measuring the behavior of large language models on a number of benchmark tasks. Using the MMLU dataset, we show that the base-rate probability (BRP) differences across answer tokens are significant and affect task performance ie. guess A if uncertain. We find that counterfactual prompting does sufficiently mitigate the BRP effect. The BRP effect is found to have a similar effect to test taking strategies employed by humans leading to the conflation of task performance and test-taking ability. We propose the Nvr-X-MMLU task, a variation of MMLU, which helps to disambiguate test-taking ability from task performance and reports the latter.
Authors: Giovanni Puccetti, Anna Rogers, Chiara Alzetta, Felice Dell'Orletta, Andrea Esuli
Abstract: Large Language Models (LLMs) are increasingly used as "content farm" models (CFMs), to generate synthetic text that could pass for real news articles. This is already happening even for languages that do not have high-quality monolingual LLMs. We show that fine-tuning Llama (v1), mostly trained on English, on as little as 40K Italian news articles, is sufficient for producing news-like texts that native speakers of Italian struggle to identify as synthetic. We investigate three LLMs and three methods of detecting synthetic texts (log-likelihood, DetectGPT, and supervised classification), finding that they all perform better than human raters, but they are all impractical in the real world (requiring either access to token likelihood information or a large dataset of CFM texts). We also explore the possibility of creating a proxy CFM: an LLM fine-tuned on a similar dataset to one used by the real "content farm". We find that even a small amount of fine-tuning data suffices for creating a successful detector, but we need to know which base LLM is used, which is a major challenge. Our results suggest that there are currently no practical methods for detecting synthetic news-like texts 'in the wild', while generating them is too easy. We highlight the urgency of more NLP research on this problem.
Authors: Seungbin Yang, ChaeHun Park, Taehee Kim, Jaegul Choo
Abstract: Recent advancements in integrating large language models (LLMs) with tools have allowed the models to interact with real-world environments. However, these \textit{tool-augmented LLMs} often encounter incomplete scenarios when users provide partial information or the necessary tools are unavailable. Recognizing and managing such scenarios is crucial for LLMs to ensure their reliability, but this exploration remains understudied. This study examines whether LLMs can identify incomplete conditions and appropriately determine when to refrain from using tools. To this end, we address a dataset by manipulating instances from two datasets by removing necessary tools or essential information for tool invocation. We confirm that most LLMs are challenged to identify the additional information required to utilize specific tools and the absence of appropriate tools. We further analyze model behaviors in different environments and compare their performance against humans. Our research can contribute to advancing reliable LLMs by addressing scenarios that commonly arise during interactions between humans and LLMs.
Authors: Yao-Ching Yu, Chun-Chih Kuo, Ziqi Ye, Yu-Cheng Chang, Yueh-Se Li
Abstract: Ensembling multiple models has always been an effective approach to push the limits of existing performance and is widely used in classification tasks by simply averaging the classification probability vectors from multiple classifiers to achieve better accuracy. However, in the thriving open-source Large Language Model (LLM) community, ensembling methods are rare and typically limited to ensembling the full-text outputs of LLMs, such as selecting the best output using a ranker, which leads to underutilization of token-level probability information. In this paper, we treat the Generation of each token by LLMs as a Classification (GaC) for ensembling. This approach fully exploits the probability information at each generation step and better prevents LLMs from producing early incorrect tokens that lead to snowballing errors. In experiments, we ensemble state-of-the-art LLMs on several benchmarks, including exams, mathematics and reasoning, and observe that our method breaks the existing community performance ceiling. Furthermore, we observed that most of the tokens in the answer are simple and do not affect the correctness of the final answer. Therefore, we also experimented with ensembling only key tokens, and the results showed better performance with lower latency across benchmarks.
Authors: Akshay Paruchuri, Jake Garrison, Shun Liao, John Hernandez, Jacob Sunshine, Tim Althoff, Xin Liu, Daniel McDuff
Abstract: Language models (LM) are capable of remarkably complex linguistic tasks; however, numerical reasoning is an area in which they frequently struggle. An important but rarely evaluated form of reasoning is understanding probability distributions. In this paper, we focus on evaluating the probabilistic reasoning capabilities of LMs using idealized and real-world statistical distributions. We perform a systematic evaluation of state-of-the-art LMs on three tasks: estimating percentiles, drawing samples, and calculating probabilities. We evaluate three ways to provide context to LMs 1) anchoring examples from within a distribution or family of distributions, 2) real-world context, 3) summary statistics on which to base a Normal approximation. Models can make inferences about distributions, and can be further aided by the incorporation of real-world context, example shots and simplified assumptions, even if these assumptions are incorrect or misspecified. To conduct this work, we developed a comprehensive benchmark distribution dataset with associated question-answer pairs that we have released publicly.
Authors: Somin Wadhwa, Silvio Amir, Byron C. Wallace
Abstract: Eliciting "chain of thought" (CoT) rationales -- sequences of token that convey a "reasoning" process -- has been shown to consistently improve LLM performance on tasks like question answering. More recent efforts have shown that such rationales can also be used for model distillation: Including CoT sequences (elicited from a large "teacher" model) in addition to target labels when fine-tuning a small student model yields (often substantial) improvements. In this work we ask: Why and how does this additional training signal help in model distillation? We perform ablations to interrogate this, and report some potentially surprising results. Specifically: (1) Placing CoT sequences after labels (rather than before) realizes consistently better downstream performance -- this means that no student "reasoning" is necessary at test time to realize gains. (2) When rationales are appended in this way, they need not be coherent reasoning sequences to yield improvements; performance increases are robust to permutations of CoT tokens, for example. In fact, (3) a small number of key tokens are sufficient to achieve improvements equivalent to those observed when full rationales are used in model distillation.
Authors: Tinh Son Luong, Thanh-Thien Le, Thang Viet Doan, Linh Ngo Van, Thien Huu Nguyen, Diep Thi-Ngoc Nguyen
Abstract: Existing toxic detection models face significant limitations, such as lack of transparency, customization, and reproducibility. These challenges stem from the closed-source nature of their training data and the paucity of explanations for their evaluation mechanism. To address these issues, we propose a dataset creation mechanism that integrates voting and chain-of-thought processes, producing a high-quality open-source dataset for toxic content detection. Our methodology ensures diverse classification metrics for each sample and includes both classification scores and explanatory reasoning for the classifications. We utilize the dataset created through our proposed mechanism to train our model, which is then compared against existing widely-used detectors. Our approach not only enhances transparency and customizability but also facilitates better fine-tuning for specific use cases. This work contributes a robust framework for developing toxic content detection models, emphasizing openness and adaptability, thus paving the way for more effective and user-specific content moderation solutions.
Authors: Ze Wang, Zekun Wu, Xin Guan, Michael Thaler, Adriano Koshiyama, Skylar Lu, Sachin Beepath, Ediz Ertekin Jr., Maria Perez-Ortiz
Abstract: The use of Large Language Models (LLMs) in hiring has led to legislative actions to protect vulnerable demographic groups. This paper presents a novel framework for benchmarking hierarchical gender hiring bias in Large Language Models (LLMs) for resume scoring, revealing significant issues of reverse gender hiring bias and overdebiasing. Our contributions are fourfold: Firstly, we introduce a new construct grounded in labour economics, legal principles, and critiques of current bias benchmarks: hiring bias can be categorized into two types: Level bias (difference in the average outcomes between demographic counterfactual groups) and Spread bias (difference in the variance of outcomes between demographic counterfactual groups); Level bias can be further subdivided into statistical bias (i.e. changing with non-demographic content) and taste-based bias (i.e. consistent regardless of non-demographic content). Secondly, the framework includes rigorous statistical and computational hiring bias metrics, such as Rank After Scoring (RAS), Rank-based Impact Ratio, Permutation Test, and Fixed Effects Model. Thirdly, we analyze gender hiring biases in ten state-of-the-art LLMs. Seven out of ten LLMs show significant biases against males in at least one industry. An industry-effect regression reveals that the healthcare industry is the most biased against males. Moreover, we found that the bias performance remains invariant with resume content for eight out of ten LLMs. This indicates that the bias performance measured in this paper might apply to other resume datasets with different resume qualities. Fourthly, we provide a user-friendly demo and resume dataset to support the adoption and practical use of the framework, which can be generalized to other social traits and tasks.
Authors: Qiancheng Xu, Yongqi Li, Heming Xia, Wenjie Li
Abstract: Tool learning aims to enhance and expand large language models' (LLMs) capabilities with external tools, which has gained significant attention recently. Current methods have shown that LLMs can effectively handle a certain amount of tools through in-context learning or fine-tuning. However, in real-world scenarios, the number of tools is typically extensive and irregularly updated, emphasizing the necessity for a dedicated tool retrieval component. Tool retrieval is nontrivial due to the following challenges: 1) complex user instructions and tool descriptions; 2) misalignment between tool retrieval and tool usage models. To address the above issues, we propose to enhance tool retrieval with iterative feedback from the large language model. Specifically, we prompt the tool usage model, i.e., the LLM, to provide feedback for the tool retriever model in multi-round, which could progressively improve the tool retriever's understanding of instructions and tools and reduce the gap between the two standalone components. We build a unified and comprehensive benchmark to evaluate tool retrieval models. The extensive experiments indicate that our proposed approach achieves advanced performance in both in-domain evaluation and out-of-domain evaluation.
Authors: Li Siyan, Teresa Shao, Zhou Yu, Julia Hirschberg
Abstract: Dialogue systems have been used as conversation partners in English learning, but few have studied whether these systems improve learning outcomes. Student passion and perseverance, or grit, has been associated with language learning success. Recent work establishes that as students perceive their English teachers to be more supportive, their grit improves. Hypothesizing that the same pattern applies to English-teaching chatbots, we create EDEN, a robust open-domain chatbot for spoken conversation practice that provides empathetic feedback. To construct EDEN, we first train a specialized spoken utterance grammar correction model and a high-quality social chit-chat conversation model. We then conduct a preliminary user study with a variety of strategies for empathetic feedback. Our experiment suggests that using adaptive empathetic feedback leads to higher perceived affective support. Furthermore, elements of perceived affective support positively correlate with student grit.
Authors: Jiwan Chung, Sungjae Lee, Minseo Kim, Seungju Han, Ashkan Yousefpour, Jack Hessel, Youngjae Yu
Abstract: Visual arguments, often used in advertising or social causes, rely on images to persuade viewers to do or believe something. Understanding these arguments requires selective vision: only specific visual stimuli within an image are relevant to the argument, and relevance can only be understood within the context of a broader argumentative structure. While visual arguments are readily appreciated by human audiences, we ask: are today's AI capable of similar understanding? We present VisArgs, a dataset of 1,611 images annotated with 5,112 visual premises (with regions), 5,574 commonsense premises, and reasoning trees connecting them into structured arguments. We propose three tasks for evaluating visual argument understanding: premise localization, premise identification, and conclusion deduction. Experiments show that 1) machines struggle to capture visual cues: GPT-4-O achieved 78.5% accuracy, while humans reached 98.0%. Models also performed 19.5% worse when distinguishing between irrelevant objects within the image compared to external objects. 2) Providing relevant visual premises improved model performance significantly.
Authors: Qihuang Zhong, Haiyun Li, Luyao Zhuang, Juhua Liu, Bo Du
Abstract: Aspect-based Sentiment Analysis (ABSA) is an important sentiment analysis task, which aims to determine the sentiment polarity towards an aspect in a sentence. Due to the expensive and limited labeled data, data generation (DG) has become the standard for improving the performance of ABSA. However, current DG methods usually have some shortcomings: 1) poor fluency and coherence, 2) lack of diversity of generated data, and 3) reliance on some existing labeled data, hindering its applications in real-world scenarios. With the advancement of large language models (LLMs), LLM-based DG has the potential to solve the above issues. Unfortunately, directly prompting LLMs struggles to generate the desired pseudo-label ABSA data, as LLMs are prone to hallucinations, leading to undesired data generation. To this end, we propose a systematic Iterative Data Generation framework, namely IDG, to boost the performance of ABSA. The core of IDG is to make full use of the powerful abilities (i.e., instruction-following, in-context learning and self-reflection) of LLMs to iteratively generate more fluent and diverse pseudo-label data, starting from an unsupervised sentence corpus. Specifically, IDG designs a novel iterative data generation mechanism and a self-reflection data filtering module to tackle the challenges of unexpected data generation caused by hallucinations. Extensive experiments on four widely-used ABSA benchmarks show that IDG brings consistent and significant performance gains among five baseline ABSA models. More encouragingly, the synthetic data generated by IDG can achieve comparable or even better performance against the manually annotated data.
Authors: Tanush Chopra, Michael Li, Jacob Haimes
Abstract: When large language models (LLMs) are asked to perform certain tasks, how can we be sure that their learned representations align with reality? We propose a domain-agnostic framework for systematically evaluating distribution shifts in LLMs decision-making processes, where they are given control of mechanisms governed by pre-defined rules. While individual LLM actions may appear consistent with expected behavior, across a large number of trials, statistically significant distribution shifts can emerge. To test this, we construct a well-defined environment with known outcome logic: blackjack. In more than 1,000 trials, we uncover statistically significant evidence suggesting behavioral misalignment in the learned representations of LLM.
Authors: Serene Lim, Mar\'ia P\'erez-Ortiz
Abstract: This paper investigates the subtle and often concealed biases present in Large Language Models (LLMs), focusing on implicit biases that may remain despite passing explicit bias tests. Implicit biases are significant because they influence the decisions made by these systems, potentially perpetuating stereotypes and discrimination, even when LLMs appear to function fairly. Traditionally, explicit bias tests or embedding-based methods are employed to detect bias, but these approaches can overlook more nuanced, implicit forms of bias. To address this, we introduce two novel psychological-inspired methodologies: the LLM Implicit Association Test (IAT) Bias and the LLM Decision Bias, designed to reveal and measure implicit biases through prompt-based and decision-making tasks. Additionally, open-ended generation tasks with thematic analysis of word generations and storytelling provide qualitative insights into the model's behavior. Our findings demonstrate that the LLM IAT Bias correlates with traditional methods and more effectively predicts downstream behaviors, as measured by the LLM Decision Bias, offering a more comprehensive framework for detecting subtle biases in AI systems. This research advances the field of AI ethics by proposing new methods to continually assess and mitigate biases in LLMs, highlighting the importance of qualitative and decision-focused evaluations to address challenges that previous approaches have not fully captured.
Authors: Ziwei Ji, Delong Chen, Etsuko Ishii, Samuel Cahyawijaya, Yejin Bang, Bryan Wilie, Pascale Fung
Abstract: The hallucination problem of Large Language Models (LLMs) significantly limits their reliability and trustworthiness. Humans have a self-awareness process that allows us to recognize what we don't know when faced with queries. Inspired by this, our paper investigates whether LLMs can estimate their own hallucination risk before response generation. We analyze the internal mechanisms of LLMs broadly both in terms of training data sources and across 15 diverse Natural Language Generation (NLG) tasks, spanning over 700 datasets. Our empirical analysis reveals two key insights: (1) LLM internal states indicate whether they have seen the query in training data or not; and (2) LLM internal states show they are likely to hallucinate or not regarding the query. Our study explores particular neurons, activation layers, and tokens that play a crucial role in the LLM perception of uncertainty and hallucination risk. By a probing estimator, we leverage LLM self-assessment, achieving an average hallucination estimation accuracy of 84.32\% at run time.
Authors: Jiahuan Cao, Dezhi Peng, Peirong Zhang, Yongxin Shi, Yang Liu, Kai Ding, Lianwen Jin
Abstract: Classical Chinese is a gateway to the rich heritage and wisdom of ancient China, yet its complexities pose formidable comprehension barriers for most modern people without specialized knowledge. While Large Language Models (LLMs) have shown remarkable capabilities in Natural Language Processing (NLP), they struggle with Classical Chinese Understanding (CCU), especially in data-demanding and knowledge-intensive tasks. In response to this dilemma, we propose \textbf{TongGu} (mean understanding ancient and modern), the first CCU-specific LLM, underpinned by three core contributions. First, we construct a two-stage instruction-tuning dataset ACCN-INS derived from rich classical Chinese corpora, aiming to unlock the full CCU potential of LLMs. Second, we propose Redundancy-Aware Tuning (RAT) to prevent catastrophic forgetting, enabling TongGu to acquire new capabilities while preserving its foundational knowledge. Third, we present a CCU Retrieval-Augmented Generation (CCU-RAG) technique to reduce hallucinations based on knowledge-grounding. Extensive experiments across 24 diverse CCU tasks validate TongGu's superior ability, underscoring the effectiveness of RAT and CCU-RAG. The model and dataset are available at \url{https://github.com/SCUT-DLVCLab/TongGu-LLM}.
Authors: Jesse Roberts, Kyle Moore, Thao Pham, Oseremhen Ewaleifoh, Doug Fisher
Abstract: This paper evaluates whether large language models (LLMs) exhibit cognitive fan effects, similar to those discovered by Anderson in humans, after being pre-trained on human textual data. We conduct two sets of in-context recall experiments designed to elicit fan effects. Consistent with human results, we find that LLM recall uncertainty, measured via token probability, is influenced by the fan effect. Our results show that removing uncertainty disrupts the observed effect. The experiments suggest the fan effect is consistent whether the fan value is induced in-context or in the pre-training data. Finally, these findings provide in-silico evidence that fan effects and typicality are expressions of the same phenomena.
Authors: Hyungjun Yoon, Biniyam Aschalew Tolera, Taesik Gong, Kimin Lee, Sung-Ju Lee
Abstract: Large language models (LLMs) have demonstrated exceptional abilities across various domains. However, utilizing LLMs for ubiquitous sensing applications remains challenging as existing text-prompt methods show significant performance degradation when handling long sensor data sequences. We propose a visual prompting approach for sensor data using multimodal LLMs (MLLMs). We design a visual prompt that directs MLLMs to utilize visualized sensor data alongside the target sensory task descriptions. Additionally, we introduce a visualization generator that automates the creation of optimal visualizations tailored to a given sensory task, eliminating the need for prior task-specific knowledge. We evaluated our approach on nine sensory tasks involving four sensing modalities, achieving an average of 10% higher accuracy than text-based prompts and reducing token costs by 15.8 times. Our findings highlight the effectiveness and cost-efficiency of visual prompts with MLLMs for various sensory tasks. The source code is available at https://github.com/diamond264/ByMyEyes.
Authors: Cheng-Kuang Wu, Zhi Rui Tam, Chao-Chung Wu, Chieh-Yen Lin, Hung-yi Lee, Yun-Nung Chen
Abstract: This study explores the proactive ability of LLMs to seek user support. We propose metrics to evaluate the trade-off between performance improvements and user burden, and investigate whether LLMs can determine when to request help under varying information availability. Our experiments show that without external feedback, many LLMs struggle to recognize their need for user support. The findings highlight the importance of external signals and provide insights for future research on improving support-seeking strategies. Source code: https://github.com/appier-research/i-need-help
Authors: Jingwei Zhu, Minghuan Tan, Min Yang, Ruixue Li, Hamid Alinejad-Rokny
Abstract: The rapid progress in Large Language Models (LLMs) has prompted the creation of numerous benchmarks to evaluate their capabilities.This study focuses on the Comprehensive Medical Benchmark in Chinese (CMB), showcasing how dataset diversity and distribution in supervised fine-tuning (SFT) may enhance LLM performance.Remarkably, We successfully trained a smaller base model to achieve scores comparable to larger models, indicating that a diverse and well-distributed dataset can optimize performance regardless of model size.This study suggests that even smaller models may reach high performance levels with carefully curated and varied datasets. By integrating a wide range of instructional content, our approach addresses potential issues such as data quality inconsistencies. Our results imply that a broader spectrum of training data may enhance a model's ability to generalize and perform effectively across different medical scenarios, highlighting the importance of dataset quality and diversity in fine-tuning processes. We open-source the model for future research at https://github.com/CAS-SIAT-XinHai/CollectiveSFT
Authors: Vil\'em Zouhar, Pinzhen Chen, Tsz Kin Lam, Nikita Moghe, Barry Haddow
Abstract: The COMET metric has blazed a trail in the machine translation community, given its strong correlation with human judgements of translation quality. Its success stems from being a modified pre-trained multilingual model finetuned for quality assessment. However, it being a machine learning model also gives rise to a new set of pitfalls that may not be widely known. We investigate these unexpected behaviours from three aspects: 1) technical: obsolete software versions and compute precision; 2) data: empty content, language mismatch, and translationese at test time as well as distribution and domain biases in training; 3) usage and reporting: multi-reference support and model referencing in the literature. All of these problems imply that COMET scores are not comparable between papers or even technical setups and we put forward our perspective on fixing each issue. Furthermore, we release the sacreCOMET package that can generate a signature for the software and model configuration as well as an appropriate citation. The goal of this work is to help the community make more sound use of the COMET metric.
Authors: Abu Ubaida Akash, Ahmed Fahmy, Amine Trabelsi
Abstract: Stance detection (SD) identifies a text's position towards a target, typically labeled as favor, against, or none. We introduce Open-Target Stance Detection (OTSD), the most realistic task where targets are neither seen during training nor provided as input. We evaluate Large Language Models (LLMs) GPT-4o, GPT-3.5, Llama-3, and Mistral, comparing their performance to the only existing work, Target-Stance Extraction (TSE), which benefits from predefined targets. Unlike TSE, OTSD removes the dependency of a predefined list, making target generation and evaluation more challenging. We also provide a metric for evaluating target quality that correlates well with human judgment. Our experiments reveal that LLMs outperform TSE in target generation when the real target is explicitly and not explicitly mentioned in the text. Likewise, for stance detection, LLMs excel in explicit cases with comparable performance in non-explicit in general.
Authors: Ray Umphrey, Jesse Roberts, Lindsey Roberts
Abstract: This study explores the potential of large language models (LLMs) for identifying and examining intertextual relationships within biblical, Koine Greek texts. By evaluating the performance of LLMs on various intertextuality scenarios the study demonstrates that these models can detect direct quotations, allusions, and echoes between texts. The LLM's ability to generate novel intertextual observations and connections highlights its potential to uncover new insights. However, the model also struggles with long query passages and the inclusion of false intertextual dependences, emphasizing the importance of expert evaluation. The expert-in-the-loop methodology presented offers a scalable approach for intertextual research into the complex web of intertextuality within and beyond the biblical corpus.
Authors: Hanlin Wang, Chak Tou Leong, Jian Wang, Wenjie Li
Abstract: Language models are exhibiting increasing capability in knowledge utilization and reasoning. However, when applied as agents in embodied environments, they often suffer from misalignment between their intrinsic knowledge and environmental knowledge, leading to infeasible actions. Traditional environment alignment methods, such as supervised learning on expert trajectories and reinforcement learning, encounter limitations in covering environmental knowledge and achieving efficient convergence, respectively. Inspired by human learning, we propose Exploration-based Error Correction Learning (E2CL), a novel framework that leverages exploration-induced errors and environmental feedback to enhance environment alignment for embodied agents. E2CL incorporates teacher-guided and teacher-free explorations to gather environmental feedback and correct erroneous actions. The agent learns to provide feedback and self-correct, thereby enhancing its adaptability to target environments. Extensive experiments in the VirtualHome environment demonstrate that E2CL-trained agents outperform those trained by baseline methods and exhibit superior self-correction capabilities.
Authors: Lihu Chen, Ga\"el Varoquaux
Abstract: Large Language Models (LLMs) have made significant progress in advancing artificial general intelligence (AGI), leading to the development of increasingly large models such as GPT-4 and LLaMA-405B. However, scaling up model sizes results in exponentially higher computational costs and energy consumption, making these models impractical for academic researchers and businesses with limited resources. At the same time, Small Models (SMs) are frequently used in practical settings, although their significance is currently underestimated. This raises important questions about the role of small models in the era of LLMs, a topic that has received limited attention in prior research. In this work, we systematically examine the relationship between LLMs and SMs from two key perspectives: Collaboration and Competition. We hope this survey provides valuable insights for practitioners, fostering a deeper understanding of the contribution of small models and promoting more efficient use of computational resources. The code is available at https://github.com/tigerchen52/role_of_small_models
Authors: Hui Yi Leong, Yi Fan Gao, Ji Shuai, Yang Zhang, Uktu Pamuksuz
Abstract: Scientific research indicates that for every hour spent in direct patient care, physicians spend nearly two additional hours on administrative tasks, particularly on electronic health records (EHRs) and desk work. This excessive administrative burden not only reduces the time available for patient care but also contributes to physician burnout and inefficiencies in healthcare delivery. To address these challenges, this study introduces MediGen, a fine-tuned large language model (LLM) designed to automate the generation of medical reports from medical dialogues. By leveraging state-of-the-art methodologies for fine-tuning open-source pretrained models, including LLaMA3-8B, MediGen achieves high accuracy in transcribing and summarizing clinical interactions. The fine-tuned LLaMA3-8B model demonstrated promising results, achieving a ROUGE score of 58% and a BERTScore-F1 of 72%, indicating its effectiveness in generating accurate and clinically relevant medical reports. These findings suggest that MediGen has the potential to significantly reduce the administrative workload on physicians, improving both healthcare efficiency and physician well-being.
Authors: Yanjun Lyu, Zihao Wu, Lu Zhang, Jing Zhang, Yiwei Li, Wei Ruan, Zhengliang Liu, Xiaowei Yu, Chao Cao, Tong Chen, Minheng Chen, Yan Zhuang, Xiang Li, Rongjie Liu, Chao Huang, Wentao Li, Tianming Liu, Dajiang Zhu
Abstract: Pre-trained large language models(LLMs) have attracted increasing attention in biomedical domains due to their success in natural language processing. However, the complex traits and heterogeneity of multi-sources genomics data pose significant challenges when adapting these models to the bioinformatics and biomedical field. To address these challenges, we present GP-GPT, the first specialized large language model for genetic-phenotype knowledge representation and genomics relation analysis. Our model is fine-tuned in two stages on a comprehensive corpus composed of over 3,000,000 terms in genomics, proteomics, and medical genetics, derived from multiple large-scale validated datasets and scientific publications. GP-GPT demonstrates proficiency in accurately retrieving medical genetics information and performing common genomics analysis tasks, such as genomics information retrieval and relationship determination. Comparative experiments across domain-specific tasks reveal that GP-GPT outperforms state-of-the-art LLMs, including Llama2, Llama3 and GPT-4. These results highlight GP-GPT's potential to enhance genetic disease relation research and facilitate accurate and efficient analysis in the fields of genomics and medical genetics. Our investigation demonstrated the subtle changes of bio-factor entities' representations in the GP-GPT, which suggested the opportunities for the application of LLMs to advancing gene-phenotype research.
Authors: Dingjie Song, Wenjun Wang, Shunian Chen, Xidong Wang, Michael Guan, Benyou Wang
Abstract: The rapid advancement of Multimodal Large Language Models (MLLMs) has led to remarkable performances across various domains. However, this progress is accompanied by a substantial surge in the resource consumption of these models. We address this pressing issue by introducing a new approach, Token Reduction using CLIP Metric (TRIM), aimed at improving the efficiency of MLLMs without sacrificing their performance. Inspired by human attention patterns in Visual Question Answering (VQA) tasks, TRIM presents a fresh perspective on the selection and reduction of image tokens. The TRIM method has been extensively tested across 12 datasets, and the results demonstrate a significant reduction in computational overhead while maintaining a consistent level of performance. This research marks a critical stride in efficient MLLM development, promoting greater accessibility and sustainability of high-performing models.
Authors: Peifeng Wang, Austin Xu, Yilun Zhou, Caiming Xiong, Shafiq Joty
Abstract: Auto-evaluation is crucial for assessing response quality and offering feedback for model development. Recent studies have explored training large language models (LLMs) as generative judges to evaluate and critique other models' outputs. In this work, we investigate the idea of learning from both positive and negative data with preference optimization to enhance the evaluation capabilities of LLM judges across an array of different use cases. We achieve this by employing three approaches to collect the preference pairs for different use cases, each aimed at improving our generative judge from a different perspective. Our comprehensive study over a wide range of benchmarks demonstrates the effectiveness of our method. In particular, our generative judge achieves the best performance on 10 out of 13 benchmarks, outperforming strong baselines like GPT-4o and specialized judge models. Further analysis show that our judge model robustly counters inherent biases such as position and length bias, flexibly adapts to any evaluation protocol specified by practitioners, and provides helpful language feedback for improving downstream generator models.
Authors: Sihui Yang, Keping Bi, Wanqing Cui, Jiafeng Guo, Xueqi Cheng
Abstract: Non-Factoid (NF) Question Answering (QA) is challenging to evaluate due to diverse potential answers and no objective criterion. The commonly used automatic evaluation metrics like ROUGE or BERTScore cannot accurately measure semantic similarities or answers from different perspectives. Recently, Large Language Models (LLMs) have been resorted to for NFQA evaluation due to their compelling performance on various NLP tasks. Common approaches include pointwise scoring of each candidate answer and pairwise comparisons between answers. Inspired by the evolution from pointwise to pairwise to listwise in learning-to-rank methods, we propose a novel listwise NFQA evaluation approach, that utilizes LLMs to rank candidate answers in a list of reference answers sorted by descending quality. Moreover, for NF questions that do not have multi-grade or any golden answers, we leverage LLMs to generate the reference answer list of various quality to facilitate the listwise evaluation. Extensive experimental results on three NFQA datasets, i.e., ANTIQUE, the TREC-DL-NF, and WebGLM show that our method has significantly higher correlations with human annotations compared to automatic scores and common pointwise and pairwise approaches.
Authors: Weichao Zhang, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke, Yixing Fan, Xueqi Cheng
Abstract: As the scale of training corpora for large language models (LLMs) grows, model developers become increasingly reluctant to disclose details on their data. This lack of transparency poses challenges to scientific evaluation and ethical deployment. Recently, pretraining data detection approaches, which infer whether a given text was part of an LLM's training data through black-box access, have been explored. The Min-K% Prob method, which has achieved state-of-the-art results, assumes that a non-training example tends to contain a few outlier words with low token probabilities. However, the effectiveness may be limited as it tends to misclassify non-training texts that contain many common words with high probabilities predicted by LLMs. To address this issue, we introduce a divergence-based calibration method, inspired by the divergence-from-randomness concept, to calibrate token probabilities for pretraining data detection. We compute the cross-entropy (i.e., the divergence) between the token probability distribution and the token frequency distribution to derive a detection score.We have developed a Chinese-language benchmark, PatentMIA, to assess the performance of detection approaches for LLMs on Chinese text. Experimental results on English-language benchmarks and PatentMIA demonstrate that our proposed method significantly outperforms existing methods.
Authors: Yuanchang Luo, Zhanglin Wu, Daimeng Wei, Hengchao Shang, Zongyao Li, Jiaxin Guo, Zhiqiang Rao, Shaojun Li, Jinlong Yang, Yuhao Xie, Jiawei Zheng Bin Wei, Hao Yang
Abstract: This article introduces the submission status of the Translation into Low-Resource Languages of Spain task at (WMT 2024) by Huawei Translation Service Center (HW-TSC). We participated in three translation tasks: spanish to aragonese (es-arg), spanish to aranese (es-arn), and spanish to asturian (es-ast). For these three translation tasks, we use training strategies such as multilingual transfer, regularized dropout, forward translation and back translation, labse denoising, transduction ensemble learning and other strategies to neural machine translation (NMT) model based on training deep transformer-big architecture. By using these enhancement strategies, our submission achieved a competitive result in the final evaluation.
Authors: Ioannis Panagiotopoulos, Giorgos Filandrianos, Maria Lymperaiou, Giorgos Stamou
Abstract: Riddle-solving requires advanced reasoning skills, pushing LLMs to engage in abstract thinking and creative problem-solving, often revealing limitations in their cognitive abilities. In this paper, we examine the riddle-solving capabilities of LLMs using a multiple-choice format, exploring how different prompting techniques impact performance on riddles that demand diverse reasoning skills. To enhance results, we introduce RISCORE (RIddle Solving with COntext REcontruciton) a novel fully automated prompting method that generates and utilizes contextually reconstructed sentence-based puzzles in conjunction with the original examples to create few-shot exemplars. Our experiments demonstrate that RISCORE significantly improves the performance of language models in both vertical and lateral thinking tasks, surpassing traditional exemplar selection strategies across a variety of few-shot settings.
Authors: Uri Berger, Edoardo M. Ponti
Abstract: Do speakers of different languages talk differently about what they see? Behavioural and cognitive studies report cultural effects on perception; however, these are mostly limited in scope and hard to replicate. In this work, we conduct the first large-scale empirical study of cross-lingual variation in image descriptions. Using a multimodal dataset with 31 languages and images from diverse locations, we develop a method to accurately identify entities mentioned in captions and present in the images, then measure how they vary across languages. Our analysis reveals that pairs of languages that are geographically or genetically closer tend to mention the same entities more frequently. We also identify entity categories whose saliency is universally high (such as animate beings), low (clothing accessories) or displaying high variance across languages (landscape). In a case study, we measure the differences in a specific language pair (e.g., Japanese mentions clothing far more frequently than English). Furthermore, our method corroborates previous small-scale studies, including 1) Rosch et al. (1976)'s theory of basic-level categories, demonstrating a preference for entities that are neither too generic nor too specific, and 2) Miyamoto et al. (2006)'s hypothesis that environments afford patterns of perception, such as entity counts. Overall, our work reveals the presence of both universal and culture-specific patterns in entity mentions.
Authors: Thong Nguyen, Truc-My Nguyen
Abstract: Counterfactual statements, which describe events that did not or cannot take place, are beneficial to numerous NLP applications. Hence, we consider the problem of counterfactual detection (CFD) and seek to enhance the CFD models. Previous models are reliant on clue phrases to predict counterfactuality, so they suffer from significant performance drop when clue phrase hints do not exist during testing. Moreover, these models tend to predict non-counterfactuals over counterfactuals. To address these issues, we propose to integrate neural topic model into the CFD model to capture the global semantics of the input statement. We continue to causally intervene the hidden representations of the CFD model to balance the effect of the class labels. Extensive experiments show that our approach outperforms previous state-of-the-art CFD and bias-resolving methods in both the CFD and other bias-sensitive tasks.
Authors: Siyue Zhang, Anh Tuan Luu, Chen Zhao
Abstract: Text-to-SQL parsing and end-to-end question answering (E2E TQA) are two main approaches for Table-based Question Answering task. Despite success on multiple benchmarks, they have yet to be compared and their synergy remains unexplored. In this paper, we identify different strengths and weaknesses through evaluating state-of-the-art models on benchmark datasets: Text-to-SQL demonstrates superiority in handling questions involving arithmetic operations and long tables; E2E TQA excels in addressing ambiguous questions, non-standard table schema, and complex table contents. To combine both strengths, we propose a Synergistic Table-based Question Answering approach that integrate different models via answer selection, which is agnostic to any model types. Further experiments validate that ensembling models by either feature-based or LLM-based answer selector significantly improves the performance over individual models.
Authors: Wanqi Yang, Yanda Li, Meng Fang, Ling Chen
Abstract: Time-Sensitive Question Answering (TSQA) demands the effective utilization of specific temporal contexts, encompassing multiple time-evolving facts, to address time-sensitive questions. This necessitates not only the parsing of temporal information within questions but also the identification and understanding of time-evolving facts to generate accurate answers. However, current large language models still have limited sensitivity to temporal information and their inadequate temporal reasoning capabilities. In this paper, we propose a novel framework that enhances temporal awareness and reasoning through Temporal Information-Aware Embedding and Granular Contrastive Reinforcement Learning. Experimental results on four TSQA datasets demonstrate that our framework significantly outperforms existing LLMs in TSQA tasks, marking a step forward in bridging the performance gap between machine and human temporal understanding and reasoning.
Authors: Shengwei Tian, Lifeng Han, Erick Mendez Guzman, Goran Nenadic
Abstract: With the rapid growth of the Natural Language Processing (NLP) field, a vast variety of Large Language Models (LLMs) continue to emerge for diverse NLP tasks. As an increasing number of papers are presented, researchers and developers face the challenge of information overload. Thus, it is particularly important to develop a system that can automatically extract and organise key information about LLMs from academic papers (\textbf{LLM model card}). This work is to develop such a pioneer system by using Named Entity Recognition (\textbf{NER}) and Relation Extraction (\textbf{RE}) methods that automatically extract key information about large language models from the papers, helping researchers to efficiently access information about LLMs. These features include model \textit{licence}, model \textit{name}, and model \textit{application}. With these features, we can form a model card for each paper. \textbf{Data-contribution} wise, 106 academic papers were processed by defining three dictionaries - LLMs name, licence, and application. 11,051 sentences were extracted through dictionary lookup, and the dataset was constructed through manual review of the final selection of 129 sentences that have a link between the name and the licence, and 106 sentences that have a link between the model name and the application. Data and code in \textsc{LLM-Card} is openly hosted at \url{https://github.com/shengwei-tian/dependency-parser-visualization}
URLs: https://github.com/shengwei-tian/dependency-parser-visualization
Authors: Linzhuang Sun, Hao Liang, Jingxuan Wei, Bihui Yu, Conghui He, Zenan Zhou, Wentao Zhang
Abstract: Large Language Models (LLMs) have exhibited exceptional performance across a broad range of tasks and domains. However, they still encounter difficulties in solving mathematical problems due to the rigorous and logical nature of mathematics. Previous studies have employed techniques such as supervised fine-tuning (SFT), prompt engineering, and search-based methods to improve the mathematical problem-solving abilities of LLMs. Despite these efforts, their performance remains suboptimal and demands substantial computational resources. To address this issue, we propose a novel approach, BEATS, to enhance mathematical problem-solving abilities. Our method leverages newly designed prompts that guide the model to iteratively rewrite, advance by one step, and generate answers based on previous steps. Additionally, we introduce a new back-verification technique that uses LLMs to validate the correctness of the generated answers. Furthermore, we employ a pruning tree search to optimize search time while achieving strong performance. Notably, our method improves Qwen2-7b-Instruct's score from 36.94 to 61.52, outperforming GPT4's 42.5 on the MATH benchmark.
Authors: Sahil Garje
Abstract: Power words are terms that evoke strong emotional responses and significantly influence readers' behavior, playing a crucial role in fields like marketing, politics, and motivational writing. This study proposes a methodology for the automated detection and analysis of power words in persuasive text using a custom lexicon created from a comprehensive dataset scraped from online sources. A specialized Python package, The Text Monger, is created and employed to identify the presence and frequency of power words within a given text. By analyzing diverse datasets, including fictional excerpts, speeches, and marketing materials,the aim is to classify and assess the impact of power words on sentiment and reader engagement. The findings provide valuable insights into the effectiveness of power words across various domains, offering practical applications for content creators, advertisers, and policymakers looking to enhance their messaging and engagement strategies.
Authors: Sergey Berezin, Reza Farahbakhsh, Noel Crespi
Abstract: We introduce a novel family of adversarial attacks that exploit the inability of language models to interpret ASCII art. To evaluate these attacks, we propose the ToxASCII benchmark and develop two custom ASCII art fonts: one leveraging special tokens and another using text-filled letter shapes. Our attacks achieve a perfect 1.0 Attack Success Rate across ten models, including OpenAI's o1-preview and LLaMA 3.1. Warning: this paper contains examples of toxic language used for research purposes.
Authors: Hang Shao, Bei Liu, Wei Wang, Xun Gong, Yanmin Qian
Abstract: As a popular multilingual and multitask pre-trained speech model, Whisper has the problem of curse of multilinguality. To enhance multilingual capabilities in small Whisper models, we propose DQ-Whisper, a novel joint distillation and quantization framework to compress Whisper for efficient inference. Firstly, we propose a novel dynamic matching distillation strategy. Then, a quantization-aware distillation framework is introduced to integrate quantization with distillation. Experimental results on various multilingual datasets show that our suggested distillation approach can effectively enhance the multilingual capabilities of small Whisper models without increasing computational costs. Up to 5.18x reduction in model size is achieved with marginal performance degradation. In addition, quantization is compatible with distillation, which can result in a higher compression rate.
Authors: Yosuke Higuchi, Tetsuji Ogawa, Tetsunori Kobayashi
Abstract: We propose to utilize an instruction-tuned large language model (LLM) for guiding the text generation process in automatic speech recognition (ASR). Modern large language models (LLMs) are adept at performing various text generation tasks through zero-shot learning, prompted with instructions designed for specific objectives. This paper explores the potential of LLMs to derive linguistic information that can facilitate text generation in end-to-end ASR models. Specifically, we instruct an LLM to correct grammatical errors in an ASR hypothesis and use the LLM-derived representations to refine the output further. The proposed model is built on the joint CTC and attention architecture, with the LLM serving as a front-end feature extractor for the decoder. The ASR hypothesis, subject to correction, is obtained from the encoder via CTC decoding and fed into the LLM along with a specific instruction. The decoder subsequently takes as input the LLM output to perform token predictions, combining acoustic information from the encoder and the powerful linguistic information provided by the LLM. Experimental results show that the proposed LLM-guided model achieves a relative gain of approximately 13\% in word error rates across major benchmarks.
Authors: Keunwoo Peter Yu, Zheyuan Zhang, Fengyuan Hu, Shane Storks, Joyce Chai
Abstract: A major reason behind the recent success of large language models (LLMs) is their \textit{in-context learning} capability, which makes it possible to rapidly adapt them to downstream text-based tasks by prompting them with a small number of relevant demonstrations. While large vision-language models (VLMs) have recently been developed for tasks requiring both text and images, they largely lack in-context learning over visual information, especially in understanding and generating text about videos. In this work, we implement \textbf{E}mergent \textbf{I}n-context \textbf{Le}arning on \textbf{V}ideos (\eilev{}), a novel training paradigm that induces in-context learning over video and text by capturing key properties of pre-training data found by prior work to be essential for in-context learning in transformers. In our experiments, we show that \eilev-trained models outperform other off-the-shelf VLMs in few-shot video narration for novel, rare actions. Furthermore, we demonstrate that these key properties of bursty distributions, skewed marginal distributions, and dynamic meaning each contribute to varying degrees to VLMs' in-context learning capability in narrating procedural videos. Our results, analysis, and \eilev{}-trained models yield numerous insights about the emergence of in-context learning over video and text, creating a foundation for future work to optimize and scale VLMs for open-domain video understanding and reasoning. Our code and demo are available at \url{https://github.com/yukw777/EILEV}.
Authors: Payam Jome Yazdian, Eric Liu, Rachel Lagasse, Hamid Mohammadi, Li Cheng, Angelica Lim
Abstract: This paper proposes MotionScript, a motion-to-text conversion algorithm and natural language representation for human body motions. MotionScript provides more detailed and accurate descriptions of human body movements compared to previous natural language methods. Most motion datasets focus on basic, well-defined actions, with limited variation in expression (e.g., sitting, walking, dribbling a ball). But for expressive actions that contain a diversity of movements in the class (e.g. being sad, dancing), or for actions outside the domain of standard motion capture datasets (e.g. stylistic walking, sign-language, interactions with animals), more specific and granular natural language descriptions are needed. Our proposed MotionScript descriptions differ from existing natural language representations in that it provides detailed descriptions in natural language rather than simple action labels or generalized captions. To the best of our knowledge, this is the first attempt at translating 3D motions to natural language descriptions without requiring training data. Our experiments demonstrate that MotionScript descriptions, when applied to text-to-motion tasks, enable large language models to generate complex, previously unseen motions. Additional examples, dataset, and code can be accessed at https://pjyazdian.github.io/MotionScript
Authors: Georgios Ioannides, Aman Chadha, Aaron Elkins
Abstract: We propose the Multi-Head Density Adaptive Attention Mechanism (DAAM), a novel probabilistic attention framework that can be used for Parameter-Efficient Fine-tuning (PEFT), and the Density Adaptive Transformer (DAT), designed to enhance information aggregation across multiple modalities, including Speech, Text, and Vision. DAAM integrates learnable mean and variance into its attention mechanism, implemented in a multi-head framework, enabling it to collectively model any probability distribution for dynamic recalibration of feature significance. This method demonstrates significant improvements, especially with highly non-stationary data, surpassing the state-of-the-art attention techniques in model performance, up to approximately +20% (abs.) in accuracy. Empirically, DAAM exhibits superior adaptability and efficacy across a diverse range of tasks, including emotion recognition in speech, image classification, and text classification, thereby establishing its robustness and versatility in handling data across multiple modalities. Furthermore, we introduce the Importance Factor, a new learning-based metric that enhances the explainability of models trained with DAAM-based methods.
Authors: Manav Singhal, Tushar Aggarwal, Abhijeet Awasthi, Nagarajan Natarajan, Aditya Kanade
Abstract: Existing evaluation benchmarks of language models of code (code LMs) focus almost exclusively on whether the LMs can generate functionally-correct code. In real-world software engineering, developers think beyond functional correctness. They have requirements on "how" a functionality should be implemented to meet overall system design objectives like efficiency, security, and maintainability. They would also trust the code LMs more if the LMs demonstrate robust understanding of such requirements. We propose a new benchmark NoFunEval to evaluate code LMs on non-functional requirements and simple classification instances for both functional and non-functional requirements. We propose a prompting method, Coding Concepts (CoCo), as a way for a developer to communicate the domain knowledge to the LMs. We conduct an extensive evaluation of 27 code LMs. Our finding is that LMs generally falter when tested on our benchmark, hinting at fundamental blindspots in their training setups. Surprisingly, even the classification accuracy on functional-correctness instances derived from the popular HumanEval benchmark is low, calling in question the depth of their comprehension and the source of their success in generating functionally-correct code in the first place. We release our benchmark and evaluation scripts publicly at https://aka.ms/NoFunEval.
Authors: Victor Zhong, Dipendra Misra, Xingdi Yuan, Marc-Alexandre C\^ot\'e
Abstract: We introduce Language Feedback Models (LFMs) that identify desirable behaviour - actions that help achieve tasks specified in the instruction - for imitation learning in instruction following. To train LFMs, we obtain feedback from Large Language Models (LLMs) on visual trajectories verbalized to language descriptions. First, by using LFMs to identify desirable behaviour to imitate, we improve in task-completion rate over strong behavioural cloning baselines on three distinct language grounding environments (Touchdown, ScienceWorld, and ALFWorld). Second, LFMs outperform using LLMs as experts to directly predict actions, when controlling for the number of LLM output tokens. Third, LFMs generalize to unseen environments, improving task-completion rate by 3.5-12.0% through one round of adaptation. Finally, LFM can be modified to provide human-interpretable feedback without performance loss, allowing human verification of desirable behaviour for imitation learning.
Authors: Eva Giboulot, Furon Teddy
Abstract: Watermarking is a technical means to dissuade malfeasant usage of Large Language Models. This paper proposes a novel watermarking scheme, so-called WaterMax, that enjoys high detectability while sustaining the quality of the generated text of the original LLM. Its new design leaves the LLM untouched (no modification of the weights, logits, temperature, or sampling technique). WaterMax balances robustness and complexity contrary to the watermarking techniques of the literature inherently provoking a trade-off between quality and robustness. Its performance is both theoretically proven and experimentally validated. It outperforms all the SotA techniques under the most complete benchmark suite. Code available at https://github.com/eva-giboulot/WaterMax.
Authors: Wenhao Wang, Yi Yang
Abstract: The arrival of Sora marks a new era for text-to-video diffusion models, bringing significant advancements in video generation and potential applications. However, Sora, along with other text-to-video diffusion models, is highly reliant on prompts, and there is no publicly available dataset that features a study of text-to-video prompts. In this paper, we introduce VidProM, the first large-scale dataset comprising 1.67 Million unique text-to-Video Prompts from real users. Additionally, this dataset includes 6.69 million videos generated by four state-of-the-art diffusion models, alongside some related data. We initially discuss the curation of this large-scale dataset, a process that is both time-consuming and costly. Subsequently, we underscore the need for a new prompt dataset specifically designed for text-to-video generation by illustrating how VidProM differs from DiffusionDB, a large-scale prompt-gallery dataset for image generation. Our extensive and diverse dataset also opens up many exciting new research areas. For instance, we suggest exploring text-to-video prompt engineering, efficient video generation, and video copy detection for diffusion models to develop better, more efficient, and safer models. The project (including the collected dataset VidProM and related code) is publicly available at https://vidprom.github.io under the CC-BY-NC 4.0 License.
Authors: Zeguan Xiao, Yan Yang, Guanhua Chen, Yun Chen
Abstract: Extensive efforts have been made before the public release of Large language models (LLMs) to align their behaviors with human values. However, even meticulously aligned LLMs remain vulnerable to malicious manipulations such as jailbreaking, leading to unintended behaviors. In this work, we propose a novel black-box jailbreak framework for automated red teaming of LLMs. We designed malicious content concealing and memory reframing with an iterative optimization algorithm to jailbreak LLMs, motivated by the research about the distractibility and over-confidence phenomenon of LLMs. Extensive experiments of jailbreaking both open-source and proprietary LLMs demonstrate the superiority of our framework in terms of effectiveness, scalability and transferability. We also evaluate the effectiveness of existing jailbreak defense methods against our attack and highlight the crucial need to develop more effective and practical defense strategies.
Authors: Run Luo, Yunshui Li, Longze Chen, Wanwei He, Ting-En Lin, Ziqiang Liu, Lei Zhang, Zikai Song, Xiaobo Xia, Tongliang Liu, Min Yang, Binyuan Hui
Abstract: The development of large language models (LLMs) has significantly advanced the emergence of large multimodal models (LMMs). While LMMs have achieved tremendous success by promoting the synergy between multimodal comprehension and creation, they often face challenges when confronted with out-of-distribution data, such as which can hardly distinguish orientation, quantity, color, structure, etc. This is primarily due to their reliance on image encoders trained to encode images into task-relevant features, which may lead them to disregard irrelevant details. Delving into the modeling capabilities of diffusion models for images naturally prompts the question: Can diffusion models serve as the eyes of large language models for image perception? In this paper, we propose DEEM, a simple but effective approach that utilizes the generative feedback of diffusion models to align the semantic distributions of the image encoder. This addresses the drawbacks of previous methods that solely relied on image encoders like CLIP-ViT, thereby enhancing the model's resilience against out-of-distribution samples and reducing visual hallucinations. Importantly, this is achieved without requiring additional training modules and with fewer training parameters. We extensively evaluated DEEM on both our newly constructed RobustVQA benchmark and other well-known benchmarks, POPE and MMVP, for visual hallucination and perception. In particular, DEEM improves LMM's visual perception performance to a large extent (e.g., 4% higher on RobustVQA, 6.5% higher on MMVP and 12.8 % higher on POPE ). Compared to the state-of-the-art interleaved content generation models, DEEM exhibits enhanced robustness and a superior capacity to alleviate model hallucinations while utilizing fewer trainable parameters, less pre-training data (10%), and a smaller base model size.
Authors: Angelica Chen, Sadhika Malladi, Lily H. Zhang, Xinyi Chen, Qiuyi Zhang, Rajesh Ranganath, Kyunghyun Cho
Abstract: Preference learning algorithms (e.g., RLHF and DPO) are frequently used to steer LLMs to produce generations that are more preferred by humans, but our understanding of their inner workings is still limited. In this work, we study the conventional wisdom that preference learning trains models to assign higher likelihoods to more preferred outputs than less preferred outputs, measured via $\textit{ranking accuracy}$. Surprisingly, we find that most state-of-the-art preference-tuned models achieve a ranking accuracy of less than 60% on common preference datasets. We furthermore derive the $\textit{idealized ranking accuracy}$ that a preference-tuned LLM would achieve if it optimized the DPO or RLHF objective perfectly. We demonstrate that existing models exhibit a significant $\textit{alignment gap}$ -- $\textit{i.e.}$, a gap between the observed and idealized ranking accuracies. We attribute this discrepancy to the DPO objective, which is empirically and theoretically ill-suited to fix even mild ranking errors in the reference model, and derive a simple and efficient formula for quantifying the difficulty of learning a given preference datapoint. Finally, we demonstrate that ranking accuracy strongly correlates with the empirically popular win rate metric when the model is close to the reference model used in the objective, shedding further light on the differences between on-policy (e.g., RLHF) and off-policy (e.g., DPO) preference learning algorithms.
Authors: Yibin Wang, Haizhou Shi, Ligong Han, Dimitris Metaxas, Hao Wang
Abstract: Large Language Models (LLMs) often suffer from overconfidence during inference, particularly when adapted to downstream domain-specific tasks with limited data. Previous work addresses this issue by employing approximate Bayesian estimation after the LLMs are trained, enabling them to quantify uncertainty. However, such post-training approaches' performance is severely limited by the parameters learned during training. In this paper, we go beyond post-training Bayesianization and propose Bayesian Low-Rank Adaptation by Backpropagation (BLoB), an algorithm that continuously and jointly adjusts both the mean and covariance of LLM parameters throughout the whole fine-tuning process. Our empirical results verify the effectiveness of BLoB in terms of generalization and uncertainty estimation, when evaluated on both in-distribution and out-of-distribution data.
Authors: Junying Chen, Chi Gui, Ruyi Ouyang, Anningzhe Gao, Shunian Chen, Guiming Hardy Chen, Xidong Wang, Ruifei Zhang, Zhenyang Cai, Ke Ji, Guangjun Yu, Xiang Wan, Benyou Wang
Abstract: The rapid development of multimodal large language models (MLLMs), such as GPT-4V, has led to significant advancements. However, these models still face challenges in medical multimodal capabilities due to limitations in the quantity and quality of medical vision-text data, stemming from data privacy concerns and high annotation costs. While pioneering approaches utilize PubMed's large-scale, de-identified medical image-text pairs to address these limitations, they still fall short due to inherent data noise. To tackle this, we refined medical image-text pairs from PubMed and employed MLLMs (GPT-4V) in an 'unblinded' capacity to denoise and reformat the data, resulting in the creation of the PubMedVision dataset with 1.3 million medical VQA samples. Our validation demonstrates that: (1) PubMedVision can significantly enhance the medical multimodal capabilities of current MLLMs, showing significant improvement in benchmarks including the MMMU Health & Medicine track; (2) manual checks by medical experts and empirical results validate the superior data quality of our dataset compared to other data construction methods. Using PubMedVision, we train a 34B medical MLLM HuatuoGPT-Vision, which shows superior performance in medical multimodal scenarios among open-source MLLMs.
Authors: Gihun Lee, Minchan Jeong, Yujin Kim, Hojung Jung, Jaehoon Oh, Sangmook Kim, Se-Young Yun
Abstract: While learning to align Large Language Models (LLMs) with human preferences has shown remarkable success, aligning these models to meet the diverse user preferences presents further challenges in preserving previous knowledge. This paper examines the impact of personalized preference optimization on LLMs, revealing that the extent of knowledge loss varies significantly with preference heterogeneity. Although previous approaches have utilized the KL constraint between the reference model and the policy model, we observe that they fail to maintain general knowledge and alignment when facing personalized preferences. To this end, we introduce Base-Anchored Preference Optimization (BAPO), a simple yet effective approach that utilizes the initial responses of reference model to mitigate forgetting while accommodating personalized alignment. BAPO effectively adapts to diverse user preferences while minimally affecting global knowledge or general alignment. Our experiments demonstrate the efficacy of BAPO in various setups.
Authors: Azmine Toushik Wasi, Mst Rafia Islam
Abstract: Integrating cognitive ergonomics with LLMs is crucial for improving safety, reliability, and user satisfaction in human-AI interactions. Current LLM designs often lack this integration, resulting in systems that may not fully align with human cognitive capabilities and limitations. This oversight exacerbates biases in LLM outputs and leads to suboptimal user experiences due to inconsistent application of user-centered design principles. Researchers are increasingly leveraging NLP, particularly LLMs, to model and understand human behavior across social sciences, psychology, psychiatry, health, and neuroscience. Our position paper explores the need to integrate cognitive ergonomics into LLM design, providing a comprehensive framework and practical guidelines for ethical development. By addressing these challenges, we aim to advance safer, more reliable, and ethically sound human-AI interactions.
Authors: Arsham Gholamzadeh Khoee, Yinan Yu, Robert Feldt, Andris Freimanis, Patrick Andersson Rhodin, Dhasarathy Parthasarathy
Abstract: Traditional methods for making software deployment decisions in the automotive industry typically rely on manual analysis of tabular software test data. These methods often lead to higher costs and delays in the software release cycle due to their labor-intensive nature. Large Language Models (LLMs) present a promising solution to these challenges. However, their application generally demands multiple rounds of human-driven prompt engineering, which limits their practical deployment, particularly for industrial end-users who need reliable and efficient results. In this paper, we propose GoNoGo, an LLM agent system designed to streamline automotive software deployment while meeting both functional requirements and practical industrial constraints. Unlike previous systems, GoNoGo is specifically tailored to address domain-specific and risk-sensitive systems. We evaluate GoNoGo's performance across different task difficulties using zero-shot and few-shot examples taken from industrial practice. Our results show that GoNoGo achieves a 100% success rate for tasks up to Level 2 difficulty with 3-shot examples, and maintains high performance even for more complex tasks. We find that GoNoGo effectively automates decision-making for simpler tasks, significantly reducing the need for manual intervention. In summary, GoNoGo represents an efficient and user-friendly LLM-based solution currently employed in our industrial partner's company to assist with software release decision-making, supporting more informed and timely decisions in the release process for risk-sensitive vehicle systems.
Authors: Boshko Koloski, Senja Pollak, Roberto Navigli, Bla\v{z} \v{S}krlj
Abstract: Large semantic knowledge bases are grounded in factual knowledge. However, recent approaches to dense text representations (i.e. embeddings) do not efficiently exploit these resources. Dense and robust representations of documents are essential for effectively solving downstream classification and retrieval tasks. This work demonstrates that injecting embedded information from knowledge bases can augment the performance of contemporary Large Language Model (LLM)-based representations for the task of text classification. Further, by considering automated machine learning (AutoML) with the fused representation space, we demonstrate it is possible to improve classification accuracy even if we use low-dimensional projections of the original representation space obtained via efficient matrix factorization. This result shows that significantly faster classifiers can be achieved with minimal or no loss in predictive performance, as demonstrated using five strong LLM baselines on six diverse real-life datasets. The code is freely available at \url{https://github.com/bkolosk1/bablfusion.git}.
Authors: Jon Saad-Falcon, Adrian Gamarra Lafuente, Shlok Natarajan, Nahum Maru, Hristo Todorov, Etash Guha, 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 15.1 and 11.2 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: Ruihao Gong, Yifu Ding, Zining Wang, Chengtao Lv, Xingyu Zheng, Jinyang Du, Haotong Qin, Jinyang Guo, Michele Magno, Xianglong Liu
Abstract: Large language models (LLMs) have achieved remarkable advancements in natural language processing, showcasing exceptional performance across various tasks. However, the expensive memory and computational requirements present significant challenges for their practical deployment. Low-bit quantization has emerged as a critical approach to mitigate these challenges by reducing the bit-width of model parameters, activations, and gradients, thus decreasing memory usage and computational demands. This paper presents a comprehensive survey of low-bit quantization methods tailored for LLMs, covering the fundamental principles, system implementations, and algorithmic strategies. An overview of basic concepts and new data formats specific to low-bit LLMs is first introduced, followed by a review of frameworks and systems that facilitate low-bit LLMs across various hardware platforms. Then, we categorize and analyze techniques and toolkits for efficient low-bit training and inference of LLMs. Finally, we conclude with a discussion of future trends and potential advancements of low-bit LLMs. Our systematic overview from basic, system, and algorithm perspectives can offer valuable insights and guidelines for future works to enhance the efficiency and applicability of LLMs through low-bit quantization.