Authors: S\'eamus Lankford, Haithem Afli, Andy Way
Abstract: In this study, a human evaluation is carried out on how hyperparameter settings impact the quality of Transformer-based Neural Machine Translation (NMT) for the low-resourced English--Irish pair. SentencePiece models using both Byte Pair Encoding (BPE) and unigram approaches were appraised. Variations in model architectures included modifying the number of layers, evaluating the optimal number of heads for attention and testing various regularisation techniques. The greatest performance improvement was recorded for a Transformer-optimized model with a 16k BPE subword model. Compared with a baseline Recurrent Neural Network (RNN) model, a Transformer-optimized model demonstrated a BLEU score improvement of 7.8 points. When benchmarked against Google Translate, our translation engines demonstrated significant improvements. Furthermore, a quantitative fine-grained manual evaluation was conducted which compared the performance of machine translation systems. Using the Multidimensional Quality Metrics (MQM) error taxonomy, a human evaluation of the error types generated by an RNN-based system and a Transformer-based system was explored. Our findings show the best-performing Transformer system significantly reduces both accuracy and fluency errors when compared with an RNN-based model.
Authors: S\'eamus Lankford, Haithem Afli, Andy Way
Abstract: adaptNMT streamlines all processes involved in the development and deployment of RNN and Transformer neural translation models. As an open-source application, it is designed for both technical and non-technical users who work in the field of machine translation. Built upon the widely-adopted OpenNMT ecosystem, the application is particularly useful for new entrants to the field since the setup of the development environment and creation of train, validation and test splits is greatly simplified. Graphing, embedded within the application, illustrates the progress of model training, and SentencePiece is used for creating subword segmentation models. Hyperparameter customization is facilitated through an intuitive user interface, and a single-click model development approach has been implemented. Models developed by adaptNMT can be evaluated using a range of metrics, and deployed as a translation service within the application. To support eco-friendly research in the NLP space, a green report also flags the power consumption and kgCO$_{2}$ emissions generated during model development. The application is freely available.
Authors: S\'eamus Lankford, Haithem Afli, Andy Way
Abstract: The advent of Multilingual Language Models (MLLMs) and Large Language Models has spawned innovation in many areas of natural language processing. Despite the exciting potential of this technology, its impact on developing high-quality Machine Translation (MT) outputs for low-resource languages remains relatively under-explored. Furthermore, an open-source application, dedicated to both fine-tuning MLLMs and managing the complete MT workflow for low-resources languages, remains unavailable. We aim to address these imbalances through the development of adaptMLLM, which streamlines all processes involved in the fine-tuning of MLLMs for MT. This open-source application is tailored for developers, translators, and users who are engaged in MT. An intuitive interface allows for easy customisation of hyperparameters, and the application offers a range of metrics for model evaluation and the capability to deploy models as a translation service directly within the application. As a multilingual tool, we used adaptMLLM to fine-tune models for two low-resource language pairs: English to Irish (EN$\leftrightarrow$GA) and English to Marathi (EN$\leftrightarrow$MR). Compared with baselines from the LoResMT2021 Shared Task, the adaptMLLM system demonstrated significant improvements. In the EN$\rightarrow$GA direction, an improvement of 5.2 BLEU points was observed and an increase of 40.5 BLEU points was recorded in the GA$\rightarrow$EN direction. Significant improvements in the translation performance of the EN$\leftrightarrow$MR pair were also observed notably in the MR$\rightarrow$EN direction with an increase of 21.3 BLEU points. Finally, a fine-grained human evaluation of the MLLM output on the EN$\rightarrow$GA pair was conducted using the Multidimensional Quality Metrics and Scalar Quality Metrics error taxonomies. The application and models are freely available.
Authors: Xin Lu, Yanyan Zhao, Bing Qin
Abstract: Pre-trained language models have been proven to possess strong base capabilities, which not only excel in in-distribution language modeling but also show powerful abilities in out-of-distribution language modeling, transfer learning and few-shot learning. Unlike existing work focusing on the influence of scale on base capabilities, our work examines the influence of architecture on those. Specifically, our concern is: How does architecture influence the base capabilities of pre-trained language models? In this work, we attempt to explain and reverse the decline in base capabilities caused by the architecture of FFN-Wider Transformers, seeking to provide some insights. Through analysis, we found the contribution ratio of Multi-Head Attention (a combination function) to pre-trained language modeling is a key factor affecting base capabilities. FFN-Wider Transformers reduce the contribution ratio of this combination function, leading to a decline in base capabilities. We confirmed this by experiments and proposed Combination Enhancement Architecture (CEA) to address the decline in base capabilities of such models. Significantly, we extended our explanation and CEA to Mixture of Experts (MoE) architecture Transformers, which also alleviated their decline in base capabilities to some extent, proving our work can offer useful guidance for architecture analysis, architecture improvement and architecture design.
Authors: Adil Soubki, John Murzaku, Arash Yousefi Jordehi, Peter Zeng, Magdalena Markowska, Seyed Abolghasem Mirroshandel, Owen Rambow
Abstract: Evaluating the theory of mind (ToM) capabilities of language models (LMs) has recently received much attention. However, many existing benchmarks rely on synthetic data which risks misaligning the resulting experiments with human behavior. We introduce the first ToM dataset based on naturally occurring spoken dialogs, Common-ToM, and show that LMs struggle to demonstrate ToM. We then show that integrating a simple, explicit representation of beliefs improves LM performance on Common-ToM.
Authors: Amit Das, Mostafa Rahgouy, Dongji Feng, Zheng Zhang, Tathagata Bhattacharya, Nilanjana Raychawdhary, Mary Sandage, Lauramarie Pope, Gerry Dozier, Cheryl Seals
Abstract: The widespread presence of offensive languages on social media has resulted in adverse effects on societal well-being. As a result, it has become very important to address this issue with high priority. Offensive languages exist in both explicit and implicit forms, with the latter being more challenging to detect. Current research in this domain encounters several challenges. Firstly, the existing datasets primarily rely on the collection of texts containing explicit offensive keywords, making it challenging to capture implicitly offensive contents that are devoid of these keywords. Secondly, usual methodologies tend to focus solely on textual analysis, neglecting the valuable insights that community information can provide. In this research paper, we introduce a novel dataset OffLanDat, a community based implicit offensive language dataset generated by ChatGPT containing data for 38 different target groups. Despite limitations in generating offensive texts using ChatGPT due to ethical constraints, we present a prompt-based approach that effectively generates implicit offensive languages. To ensure data quality, we evaluate our data with human. Additionally, we employ a prompt-based Zero-Shot method with ChatGPT and compare the detection results between human annotation and ChatGPT annotation. We utilize existing state-of-the-art models to see how effective they are in detecting such languages. We will make our code and dataset public for other researchers.
Authors: Krishnapriya Vishnubhotla, Adam Hammond, Graeme Hirst, Saif M. Mohammad
Abstract: Stories are rich in the emotions they exhibit in their narratives and evoke in the readers. The emotional journeys of the various characters within a story are central to their appeal. Computational analysis of the emotions of novels, however, has rarely examined the variation in the emotional trajectories of the different characters within them, instead considering the entire novel to represent a single story arc. In this work, we use character dialogue to distinguish between the emotion arcs of the narration and the various characters. We analyze the emotion arcs of the various characters in a dataset of English literary novels using the framework of Utterance Emotion Dynamics. Our findings show that the narration and the dialogue largely express disparate emotions through the course of a novel, and that the commonalities or differences in the emotional arcs of stories are more accurately captured by those associated with individual characters.
Authors: Stephanie Day, Jin K. Hwang, Tracy Arner, Danielle McNamara, Carol Connor
Abstract: The purpose of this feasibility study was to examine the potential impact of reading digital interactive e-books on essential skills that support reading comprehension with third-fifth grade students. Students read two e-Books that taught word learning and comprehension monitoring strategies in the service of learning difficult vocabulary and targeted science concepts about hurricanes. We investigated whether specific comprehension strategies including word learning and strategies that supported general reading comprehension, summarization, and question generation, show promise of effectiveness in building vocabulary knowledge and comprehension skills in the e-Books. Students were assigned to read one of three versions of each of the e-Books, each version implemented one strategy. The books employed a choose-your-adventure format with embedded comprehension questions that provided students with immediate feedback on their responses. Paired samples t-tests were run to examine pre-to-post differences in learning the targeted vocabulary and science concepts taught in both e-Books. For both e-Books, students demonstrated significant gains in word learning and on the targeted hurricane concepts. Additionally, Hierarchical Linear Modeling (HLM) revealed that no one strategy was more associated with larger gains than the other. Performance on the embedded questions in the books was also associated with greater posttest outcomes for both e-Books. This work discusses important considerations for implementation and future development of e-books that can enhance student engagement and improve reading comprehension.
Authors: Yifan Song, Da Yin, Xiang Yue, Jie Huang, Sujian Li, Bill Yuchen Lin
Abstract: Large Language Models (LLMs) have become integral components in various autonomous agent systems. In this study, we present an exploration-based trajectory optimization approach, referred to as ETO. This learning method is designed to enhance the performance of open LLM agents. Contrary to previous studies that exclusively train on successful expert trajectories, our method allows agents to learn from their exploration failures. This leads to improved performance through an iterative optimization framework. During the exploration phase, the agent interacts with the environment while completing given tasks, gathering failure trajectories to create contrastive trajectory pairs. In the subsequent training phase, the agent utilizes these trajectory preference pairs to update its policy using contrastive learning methods like DPO. This iterative cycle of exploration and training fosters continued improvement in the agents. Our experiments on three complex tasks demonstrate that ETO consistently surpasses baseline performance by a large margin. Furthermore, an examination of task-solving efficiency and potential in scenarios lacking expert trajectory underscores the effectiveness of our approach.
Authors: Yu Wang
Abstract: The pretrain-finetune paradigm represents a transformative approach in natural language processing (NLP). This paradigm distinguishes itself through the use of large pretrained language models, demonstrating remarkable efficiency in finetuning tasks, even with limited training data. This efficiency is especially beneficial for research in social sciences, where the number of annotated samples is often quite limited. Our tutorial offers a comprehensive introduction to the pretrain-finetune paradigm. We first delve into the fundamental concepts of pretraining and finetuning, followed by practical exercises using real-world applications. We demonstrate the application of the paradigm across various tasks, including multi-class classification and regression. Emphasizing its efficacy and user-friendliness, the tutorial aims to encourage broader adoption of this paradigm. To this end, we have provided open access to all our code and datasets. The tutorial is particularly valuable for quantitative researchers in psychology, offering them an insightful guide into this innovative approach.
Authors: Xiang Gao, Jiaxin Zhang, Lalla Mouatadid, Kamalika Das
Abstract: In recent years, large language models (LLMs) have become increasingly prevalent, offering remarkable text generation capabilities. However, a pressing challenge is their tendency to make confidently wrong predictions, highlighting the critical need for uncertainty quantification (UQ) in LLMs. While previous works have mainly focused on addressing aleatoric uncertainty, the full spectrum of uncertainties, including epistemic, remains inadequately explored. Motivated by this gap, we introduce a novel UQ method, sampling with perturbation for UQ (SPUQ), designed to tackle both aleatoric and epistemic uncertainties. The method entails generating a set of perturbations for LLM inputs, sampling outputs for each perturbation, and incorporating an aggregation module that generalizes the sampling uncertainty approach for text generation tasks. Through extensive experiments on various datasets, we investigated different perturbation and aggregation techniques. Our findings show a substantial improvement in model uncertainty calibration, with a reduction in Expected Calibration Error (ECE) by 50\% on average. Our findings suggest that our proposed UQ method offers promising steps toward enhancing the reliability and trustworthiness of LLMs.
Authors: Chen Zheng, Ke Sun, Hang Wu, Chenguang Xi, Xun Zhou
Abstract: In recent advancements in Conversational Large Language Models (LLMs), a concerning trend has emerged, showing that many new base LLMs experience a knowledge reduction in their foundational capabilities following Supervised Fine-Tuning (SFT). This process often leads to issues such as forgetting or a decrease in the base model's abilities. Moreover, fine-tuned models struggle to align with user preferences, inadvertently increasing the generation of toxic outputs when specifically prompted. To overcome these challenges, we adopted an innovative approach by completely bypassing SFT and directly implementing Harmless Reinforcement Learning from Human Feedback (RLHF). Our method not only preserves the base model's general capabilities but also significantly enhances its conversational abilities, while notably reducing the generation of toxic outputs. Our approach holds significant implications for fields that demand a nuanced understanding and generation of responses, such as customer service. We applied this methodology to Mistral, the most popular base model, thereby creating Mistral-Plus. Our validation across 11 general tasks demonstrates that Mistral-Plus outperforms similarly sized open-source base models and their corresponding instruct versions. Importantly, the conversational abilities of Mistral-Plus were significantly improved, indicating a substantial advancement over traditional SFT models in both safety and user preference alignment.
Authors: Xueqing Wu, Rui Zheng, Jingzhen Sha, Te-Lin Wu, Hanyu Zhou, Mohan Tang, Kai-Wei Chang, Nanyun Peng, Haoran Huang
Abstract: Data analysis is a crucial analytical process to generate in-depth studies and conclusive insights to comprehensively answer a given user query for tabular data. In this work, we aim to propose new resources and benchmarks to inspire future research on this crucial yet challenging and under-explored task. However, collecting data analysis annotations curated by experts can be prohibitively expensive. We propose to automatically generate high-quality answer annotations leveraging the code-generation capabilities of LLMs with a multi-turn prompting technique. We construct the DACO dataset, containing (1) 440 databases (of tabular data) collected from real-world scenarios, (2) ~2k query-answer pairs that can serve as weak supervision for model training, and (3) a concentrated but high-quality test set with human refined annotations that serves as our main evaluation benchmark. We train a 6B supervised fine-tuning (SFT) model on DACO dataset, and find that the SFT model learns reasonable data analysis capabilities. To further align the models with human preference, we use reinforcement learning to encourage generating analysis perceived by human as helpful, and design a set of dense rewards to propagate the sparse human preference reward to intermediate code generation steps. Our DACO-RL algorithm is evaluated by human annotators to produce more helpful answers than SFT model in 57.72% cases, validating the effectiveness of our proposed algorithm. Data and code are released at https://github.com/shirley-wu/daco
Authors: Brenda Y. Miao, Irene Y. Chen, Christopher YK Williams, Jays\'on Davidson, Augusto Garcia-Agundez, Harry Sun, Travis Zack, Atul J. Butte, Madhumita Sushil
Abstract: Recent advances in generative models, including large language models (LLMs), vision language models (VLMs), and diffusion models, have accelerated the field of natural language and image processing in medicine and marked a significant paradigm shift in how biomedical models can be developed and deployed. While these models are highly adaptable to new tasks, scaling and evaluating their usage presents new challenges not addressed in previous frameworks. In particular, the ability of these models to produce useful outputs with little to no specialized training data ("zero-" or "few-shot" approaches), as well as the open-ended nature of their outputs, necessitate the development of updated guidelines in using and evaluating these models. In response to gaps in standards and best practices for the development of clinical AI tools identified by US Executive Order 141103 and several emerging national networks for clinical AI evaluation, we begin to formalize some of these guidelines by building on the "Minimum information about clinical artificial intelligence modeling" (MI-CLAIM) checklist. The MI-CLAIM checklist, originally developed in 2020, provided a set of six steps with guidelines on the minimum information necessary to encourage transparent, reproducible research for artificial intelligence (AI) in medicine. Here, we propose modifications to the original checklist that highlight differences in training, evaluation, interpretability, and reproducibility of generative models compared to traditional AI models for clinical research. This updated checklist also seeks to clarify cohort selection reporting and adds additional items on alignment with ethical standards.
Authors: Bryan Li, Tamer Alkhouli, Daniele Bonadiman, Nikolaos Pappas, Saab Mansour
Abstract: Development of large language models (LLM) have shown progress on reasoning, though studies have been limited to English or simple reasoning tasks. We thus introduce a multilingual structured reasoning and explanation dataset, termed xSTREET, that covers four tasks across six languages. xSTREET exposes a gap in base LLM performance between English and non-English reasoning tasks. We then propose two methods to remedy this gap, building on the insight that LLMs trained on code are better reasoners. First, at training time, we augment a code dataset with multi-lingual comments using machine translation while keeping program code as-is. Second, at inference time, we bridge the gap between training and inference by employing a prompt structure that incorporates step-by-step code primitives to derive new facts and find a solution. Our methods show improved multilingual performance on xSTREET, most notably on the scientific commonsense reasoning subtask. Furthermore, the models show no regression on non-reasoning tasks, thus showing our techniques maintain general-purpose abilities.
Authors: Zefan Cai, Po-Nien Kung, Ashima Suvarna, Mingyu Derek Ma, Hritik Bansal, Baobao Chang, P. Jeffrey Brantingham, Wei Wang, Nanyun Peng
Abstract: Existing approaches on zero-shot event detection usually train models on datasets annotated with known event types, and prompt them with unseen event definitions. These approaches yield sporadic successes, yet generally fall short of expectations. In this work, we aim to improve zero-shot event detection by training models to better follow event definitions. We hypothesize that a diverse set of event types and definitions are the key for models to learn to follow event definitions while existing event extraction datasets focus on annotating many high-quality examples for a few event types. To verify our hypothesis, we construct an automatically generated Diverse Event Definition (DivED) dataset and conduct comparative studies. Our experiments reveal that a large number of event types (200) and diverse event definitions can significantly boost event extraction performance; on the other hand, the performance does not scale with over ten examples per event type. Beyond scaling, we incorporate event ontology information and hard-negative samples during training, further boosting the performance. Based on these findings, we fine-tuned a LLaMA-2-7B model on our DivED dataset, yielding performance that surpasses SOTA large language models like GPT-3.5 across three open benchmarks on zero-shot event detection.
Authors: Jinman Zhao, Xueyan Zhang
Abstract: We present a comprehensive evaluation of large language models(LLMs)' ability to reason about composition relations through a benchmark encompassing 1,500 test cases in English, designed to cover six distinct types of composition relations: Positional, Comparative, Personal, Mathematical, Identity, and Other. Acknowledging the significance of multilingual capabilities, we expanded our assessment to include translations of these cases into Chinese, Japanese, French, and Korean. Our Multilingual Composition Relation (MCR) benchmark aims at investigating the robustness and adaptability of LLMs in handling composition relation reasoning across diverse linguistic contexts.
Authors: Xiangyu Li, Xinjie Shen, Yawen Zeng, Xiaofen Xing, Jin Xu
Abstract: The task of stock earnings forecasting has received considerable attention due to the demand investors in real-world scenarios. However, compared with financial institutions, it is not easy for ordinary investors to mine factors and analyze news. On the other hand, although large language models in the financial field can serve users in the form of dialogue robots, it still requires users to have financial knowledge to ask reasonable questions. To serve the user experience, we aim to build an automatic system, FinReport, for ordinary investors to collect information, analyze it, and generate reports after summarizing. Specifically, our FinReport is based on financial news announcements and a multi-factor model to ensure the professionalism of the report. The FinReport consists of three modules: news factorization module, return forecasting module, risk assessment module. The news factorization module involves understanding news information and combining it with stock factors, the return forecasting module aim to analysis the impact of news on market sentiment, and the risk assessment module is adopted to control investment risk. Extensive experiments on real-world datasets have well verified the effectiveness and explainability of our proposed FinReport. Our codes and datasets are available at https://github.com/frinkleko/FinReport.
Authors: Masamune Kobayashi, Masato Mita, Mamoru Komachi
Abstract: Metrics are the foundation for automatic evaluation in grammatical error correction (GEC), with their evaluation of the metrics (meta-evaluation) relying on their correlation with human judgments. However, conventional meta-evaluations in English GEC encounter several challenges including biases caused by inconsistencies in evaluation granularity, and an outdated setup using classical systems. These problems can lead to misinterpretation of metrics and potentially hinder the applicability of GEC techniques. To address these issues, this paper proposes SEEDA, a new dataset for GEC meta-evaluation. SEEDA consists of corrections with human ratings along two different granularities: edit-based and sentence-based, covering 12 state-of-the-art systems including large language models (LLMs), and two human corrections with different focuses. The results of improved correlations by aligning the granularity in the sentence-level meta-evaluation, suggest that edit-based metrics may have been underestimated in existing studies. Furthermore, correlations of most metrics decrease when changing from classical to neural systems, indicating that traditional metrics are relatively poor at evaluating fluently corrected sentences with many edits.
Authors: Qiusi Zhan, Zhixiang Liang, Zifan Ying, Daniel Kang
Abstract: Recent work has embodied LLMs as agents, allowing them to access tools, perform actions, and interact with external content (e.g., emails or websites). However, external content introduces the risk of indirect prompt injection (IPI) attacks, where malicious instructions are embedded within the content processed by LLMs, aiming to manipulate these agents into executing detrimental actions against users. Given the potentially severe consequences of such attacks, establishing benchmarks to assess and mitigate these risks is imperative. In this work, we introduce InjecAgent, a benchmark designed to assess the vulnerability of tool-integrated LLM agents to IPI attacks. InjecAgent comprises 1,054 test cases covering 17 different user tools and 62 attacker tools. We categorize attack intentions into two primary types: direct harm to users and exfiltration of private data. We evaluate 30 different LLM agents and show that agents are vulnerable to IPI attacks, with ReAct-prompted GPT-4 vulnerable to attacks 24% of the time. Further investigation into an enhanced setting, where the attacker instructions are reinforced with a hacking prompt, shows additional increases in success rates, nearly doubling the attack success rate on the ReAct-prompted GPT-4. Our findings raise questions about the widespread deployment of LLM Agents. Our benchmark is available at https://github.com/uiuc-kang-lab/InjecAgent.
Authors: Congzhi Zhang, Linhai Zhang, Deyu Zhou
Abstract: Conventional multi-hop fact verification models are prone to rely on spurious correlations from the annotation artifacts, leading to an obvious performance decline on unbiased datasets. Among the various debiasing works, the causal inference-based methods become popular by performing theoretically guaranteed debiasing such as casual intervention or counterfactual reasoning. However, existing causal inference-based debiasing methods, which mainly formulate fact verification as a single-hop reasoning task to tackle shallow bias patterns, cannot deal with the complicated bias patterns hidden in multiple hops of evidence. To address the challenge, we propose Causal Walk, a novel method for debiasing multi-hop fact verification from a causal perspective with front-door adjustment. Specifically, in the structural causal model, the reasoning path between the treatment (the input claim-evidence graph) and the outcome (the veracity label) is introduced as the mediator to block the confounder. With the front-door adjustment, the causal effect between the treatment and the outcome is decomposed into the causal effect between the treatment and the mediator, which is estimated by applying the idea of random walk, and the causal effect between the mediator and the outcome, which is estimated with normalized weighted geometric mean approximation. To investigate the effectiveness of the proposed method, an adversarial multi-hop fact verification dataset and a symmetric multi-hop fact verification dataset are proposed with the help of the large language model. Experimental results show that Causal Walk outperforms some previous debiasing methods on both existing datasets and the newly constructed datasets. Code and data will be released at https://github.com/zcccccz/CausalWalk.
Authors: Chan-Jan Hsu, Chang-Le Liu, Feng-Ting Liao, Po-Chun Hsu, Yi-Chang Chen, Da-Shan Shiu
Abstract: Breeze-7B is an open-source language model based on Mistral-7B, designed to address the need for improved language comprehension and chatbot-oriented capabilities in Traditional Chinese. This technical report provides an overview of the additional pretraining, finetuning, and evaluation stages for the Breeze-7B model. The Breeze-7B family of base and chat models exhibits good performance on language comprehension and chatbot-oriented tasks, reaching the top in several benchmarks among models comparable in its complexity class.
Authors: Jiwen Zhang, Jihao Wu, Yihua Teng, Minghui Liao, Nuo Xu, Xiao Xiao, Zhongyu Wei, Duyu Tang
Abstract: Large language model (LLM) leads to a surge of autonomous GUI agents for smartphone, which completes a task triggered by natural language through predicting a sequence of actions of API. Even though the task highly relies on past actions and visual observations, existing studies typical consider little semantic information carried out by intermediate screenshots and screen operations. To address this, this work presents Chain-of-Action-Thought (dubbed CoAT), which takes the description of the previous actions, the current screen, and more importantly the action thinking of what actions should be performed and the outcomes led by the chosen action. We demonstrate that, in a zero-shot setting upon an off-the-shell LLM, CoAT significantly improves the goal progress compared to standard context modeling. To further facilitate the research in this line, we construct a benchmark Android-In-The-Zoo (AitZ), which contains 18,643 screen-action pairs together with chain-of-action-thought annotations. Experiments show that fine-tuning a 200M model on our AitZ dataset achieves on par performance with CogAgent-Chat-18B.
Authors: Sang T. Truong, Duc Q. Nguyen, Toan Nguyen, Dong D. Le, Nhi N. Truong, Tho Quan, Sanmi Koyejo
Abstract: Recent advancements in large language models (LLMs) have underscored their importance in the evolution of artificial intelligence. However, despite extensive pretraining on multilingual datasets, available open-sourced LLMs exhibit limited effectiveness in processing Vietnamese. The challenge is exacerbated by the absence of systematic benchmark datasets and metrics tailored for Vietnamese LLM evaluation. To mitigate these issues, we have finetuned LLMs specifically for Vietnamese and developed a comprehensive evaluation framework encompassing 10 common tasks and 31 metrics. Our evaluation results reveal that the fine-tuned LLMs exhibit enhanced comprehension and generative capabilities in Vietnamese. Moreover, our analysis indicates that models with more parameters can introduce more biases and uncalibrated outputs and the key factor influencing LLM performance is the quality of the training or fine-tuning datasets. These insights underscore the significance of meticulous fine-tuning with high-quality datasets in enhancing LLM performance.
Authors: Mengyi Huang, Meng Xiao, Ludi Wang, Yi Du
Abstract: Continuous Relation Extraction (CRE) aims to incrementally learn relation knowledge from a non-stationary stream of data. Since the introduction of new relational tasks can overshadow previously learned information, catastrophic forgetting becomes a significant challenge in this domain. Current replay-based training paradigms prioritize all data uniformly and train memory samples through multiple rounds, which would result in overfitting old tasks and pronounced bias towards new tasks because of the imbalances of the replay set. To handle the problem, we introduce the DecouPled CRE (DP-CRE) framework that decouples the process of prior information preservation and new knowledge acquisition. This framework examines alterations in the embedding space as new relation classes emerge, distinctly managing the preservation and acquisition of knowledge. Extensive experiments show that DP-CRE significantly outperforms other CRE baselines across two datasets.
Authors: Sijie Ji, Xinzhe Zheng, Chenshu Wu
Abstract: There is an ongoing debate regarding the potential of Large Language Models (LLMs) as foundational models seamlessly integrated with Cyber-Physical Systems (CPS) for interpreting the physical world. In this paper, we carry out a case study to answer the following question: Are LLMs capable of zero-shot human activity recognition (HAR). Our study, HARGPT, presents an affirmative answer by demonstrating that LLMs can comprehend raw IMU data and perform HAR tasks in a zero-shot manner, with only appropriate prompts. HARGPT inputs raw IMU data into LLMs and utilizes the role-play and think step-by-step strategies for prompting. We benchmark HARGPT on GPT4 using two public datasets of different inter-class similarities and compare various baselines both based on traditional machine learning and state-of-the-art deep classification models. Remarkably, LLMs successfully recognize human activities from raw IMU data and consistently outperform all the baselines on both datasets. Our findings indicate that by effective prompting, LLMs can interpret raw IMU data based on their knowledge base, possessing a promising potential to analyze raw sensor data of the physical world effectively.
Authors: Congzhi Zhang, Linhai Zhang, Deyu Zhou, Guoqiang Xu
Abstract: Despite the significant achievements of existing prompting methods such as in-context learning and chain-of-thought for large language models (LLMs), they still face challenges of various biases. Traditional debiasing methods primarily focus on the model training stage, including data augmentation-based and reweight-based approaches, with the limitations of addressing the complex biases of LLMs. To address such limitations, the causal relationship behind the prompting methods is uncovered using a structural causal model, and a novel causal prompting method based on front-door adjustment is proposed to effectively mitigate the bias of LLMs. In specific, causal intervention is implemented by designing the prompts without accessing the parameters and logits of LLMs.The chain-of-thoughts generated by LLMs are employed as the mediator variable and the causal effect between the input prompt and the output answers is calculated through front-door adjustment to mitigate model biases. Moreover, to obtain the representation of the samples precisely and estimate the causal effect more accurately, contrastive learning is used to fine-tune the encoder of the samples by aligning the space of the encoder with the LLM. Experimental results show that the proposed causal prompting approach achieves excellent performance on 3 natural language processing datasets on both open-source and closed-source LLMs.
Authors: Zhonghai Wang, Jie Jiang, Yibing Zhan, Bohao Zhou, Yanhong Li, Chong Zhang, Liang Ding, Hua Jin, Jun Peng, Xu Lin, Weifeng Liu
Abstract: Medical large language models (LLMs) have gained popularity recently due to their significant practical utility. However, most existing research focuses on general medicine, and there is a need for in-depth study of LLMs in specific fields like anesthesiology. To fill the gap, we introduce Hypnos, a Chinese Anesthesia model built upon existing LLMs, e.g., Llama. Hypnos' contributions have three aspects: 1) The data, such as utilizing Self-Instruct, acquired from current LLMs likely includes inaccuracies. Hypnos implements a cross-filtering strategy to improve the data quality. This strategy involves using one LLM to assess the quality of the generated data from another LLM and filtering out the data with low quality. 2) Hypnos employs a general-to-specific training strategy that starts by fine-tuning LLMs using the general medicine data and subsequently improving the fine-tuned LLMs using data specifically from Anesthesiology. The general medical data supplement the medical expertise in Anesthesiology and enhance the effectiveness of Hypnos' generation. 3) We introduce a standardized benchmark for evaluating medical LLM in Anesthesiology. Our benchmark includes both publicly available instances from the Internet and privately obtained cases from the Hospital. Hypnos outperforms other medical LLMs in anesthesiology in metrics, GPT-4, and human evaluation on the benchmark dataset.
Authors: Rui Wang, Fei Mi, Yi Chen, Boyang Xue, Hongru Wang, Qi Zhu, Kam-Fai Wong, Ruifeng Xu
Abstract: The growing interest in Large Language Models (LLMs) for specialized applications has revealed a significant challenge: when tailored to specific domains, LLMs tend to experience catastrophic forgetting, compromising their general capabilities and leading to a suboptimal user experience. Additionally, crafting a versatile model for multiple domains simultaneously often results in a decline in overall performance due to confusion between domains. In response to these issues, we present the RolE Prompting Guided Multi-Domain Adaptation (REGA) strategy. This novel approach effectively manages multi-domain LLM adaptation through three key components: 1) Self-Distillation constructs and replays general-domain exemplars to alleviate catastrophic forgetting. 2) Role Prompting assigns a central prompt to the general domain and a unique role prompt to each specific domain to minimize inter-domain confusion during training. 3) Role Integration reuses and integrates a small portion of domain-specific data to the general-domain data, which are trained under the guidance of the central prompt. The central prompt is used for a streamlined inference process, removing the necessity to switch prompts for different domains. Empirical results demonstrate that REGA effectively alleviates catastrophic forgetting and inter-domain confusion. This leads to improved domain-specific performance compared to standard fine-tuned models, while still preserving robust general capabilities.
Authors: Bo Wang, Tianxiang Sun, Hang Yan, Siyin Wang, Qingyuan Cheng, Xipeng Qiu
Abstract: The exploration of whether agents can align with their environment without relying on human-labeled data presents an intriguing research topic. Drawing inspiration from the alignment process observed in intelligent organisms, where declarative memory plays a pivotal role in summarizing past experiences, we propose a novel learning framework. The agents adeptly distill insights from past experiences, refining and updating existing notes to enhance their performance in the environment. This entire process transpires within the memory components and is implemented through natural language, so we character this framework as In-memory Learning. We also delve into the key features of benchmarks designed to evaluate the self-improvement process. Through systematic experiments, we demonstrate the effectiveness of our framework and provide insights into this problem.
Authors: Yaochen Zhu, Rui Xia, Jiajun Zhang
Abstract: Model merging is to combine fine-tuned models derived from multiple domains, with the intent of enhancing the model's proficiency across various domains. The principal concern is the resolution of parameter conflicts. A substantial amount of existing research remedy this issue during the merging stage, with the latest study focusing on resolving this issue throughout the pruning stage. The DARE approach has exhibited promising outcomes when applied to a simplistic fine-tuned model. However, the efficacy of this method tends to wane when employed on complex fine-tuned models that show a significant parameter bias relative to the baseline model. In this paper, we introduce a dual-stage method termed Dynamic Pruning Partition Amplification (DPPA), devised to tackle the challenge of merging complex fine-tuned models. Initially, we introduce Dynamically Pruning (DP), an improved approach based on magnitude pruning, which aim is to enhance performance at higher pruning rates. Subsequently, we propose Dynamically Partition Amplification (DPA), a rescaling strategy, is designed to dynamically amplify parameter partitions in relation to their significance levels. The experimental results show that our method maintains a mere 20% of domain-specific parameters and yet delivers a performance comparable to other methodologies that preserve up to 90% of parameters. Furthermore, our method displays outstanding performance post-pruning, leading to a significant improvement of nearly 20% performance in model merging. We make our code on Github.
Authors: Hui Huang, Yingqi Qu, Jing Liu, Muyun Yang, Tiejun Zhao
Abstract: Recently, there has been a growing trend of utilizing Large Language Model (LLM) to evaluate the quality of other LLMs. Many studies have employed proprietary close-source models, especially GPT4, as the evaluator. Alternatively, other works have fine-tuned judge models based on open-source LLMs as the evaluator. In this study, we conduct an empirical study of different judge models on their evaluation capability. Our findings indicate that although the fine-tuned judge models achieve high accuracy on in-domain test sets, even surpassing GPT4, they are inherently task-specific classifiers, and their generalizability and fairness severely underperform GPT4.
Authors: Zhengyang Tang, Xingxing Zhang, Benyou Wan, Furu Wei
Abstract: Large language models (LLMs) have demonstrated remarkable capabilities in problem-solving. However, their proficiency in solving mathematical problems remains inadequate. We propose MathScale, a simple and scalable method to create high-quality mathematical reasoning data using frontier LLMs (e.g., {\tt GPT-3.5}). Inspired by the cognitive mechanism in human mathematical learning, it first extracts topics and knowledge points from seed math questions and then build a concept graph, which is subsequently used to generate new math questions. MathScale exhibits effective scalability along the size axis of the math dataset that we generate. As a result, we create a mathematical reasoning dataset (MathScaleQA) containing two million math question-answer pairs. To evaluate mathematical reasoning abilities of LLMs comprehensively, we construct {\sc MwpBench}, a benchmark of Math Word Problems, which is a collection of ten datasets (including GSM8K and MATH) covering K-12, college, and competition level math problems. We apply MathScaleQA to fine-tune open-source LLMs (e.g., LLaMA-2 and Mistral), resulting in significantly improved capabilities in mathematical reasoning. Evaluated on {\sc MwpBench}, MathScale-7B achieves state-of-the-art performance across all datasets, surpassing its best peers of equivalent size by 42.9\% in micro average accuracy and 43.7\% in macro average accuracy, respectively.
Authors: Yakir Yehuda, Itzik Malkiel, Oren Barkan, Jonathan Weill, Royi Ronen, Noam Koenigstein
Abstract: Despite the many advances of Large Language Models (LLMs) and their unprecedented rapid evolution, their impact and integration into every facet of our daily lives is limited due to various reasons. One critical factor hindering their widespread adoption is the occurrence of hallucinations, where LLMs invent answers that sound realistic, yet drift away from factual truth. In this paper, we present a novel method for detecting hallucinations in large language models, which tackles a critical issue in the adoption of these models in various real-world scenarios. Through extensive evaluations across multiple datasets and LLMs, including Llama-2, we study the hallucination levels of various recent LLMs and demonstrate the effectiveness of our method to automatically detect them. Notably, we observe up to 62% hallucinations for Llama-2 in a specific experiment, where our method achieves a Balanced Accuracy (B-ACC) of 87%, all without relying on external knowledge.
Authors: Zhitao He, Pengfei Cao, Yubo Chen, Kang Liu, Zhiqiang Zhang, Mengshu Sun, Jun Zhao
Abstract: Event Causality Identification (ECI) refers to detect causal relations between events in texts. However, most existing studies focus on sentence-level ECI with high-resource language, leaving more challenging document-level ECI (DECI) with low-resource languages under-explored. In this paper, we propose a Heterogeneous Graph Interaction Model with Multi-granularity Contrastive Transfer Learning (GIMC) for zero-shot cross-lingual document-level ECI. Specifically, we introduce a heterogeneous graph interaction network to model the long-distance dependencies between events that are scattered over document. Then, to improve cross-lingual transferability of causal knowledge learned from source language, we propose a multi-granularity contrastive transfer learning module to align the causal representations across languages. Extensive experiments show our framework outperforms previous state-of-the-art model by 9.4% and 8.2% of average F1 score on monolingual and multilingual scenarios respectively. Notably, in multilingual scenario, our zero-shot framework even exceeds GPT-3.5 with few-shot learning by 24.3% in overall performance.
Authors: Chengguang Gan, Xuzheng He, Qinghao Zhang, Tatsunori Mori
Abstract: The Mutual Reinforcement Effect (MRE) investigates the synergistic relationship between word-level and text-level classifications in text classification tasks. It posits that the performance of both classification levels can be mutually enhanced. However, this mechanism has not been adequately demonstrated or explained in prior research. To address this gap, we employ information flow analysis to observe and substantiate the MRE theory. Our experiments on six MRE hybrid datasets revealed the presence of MRE in the model and its impact. Additionally, we conducted fine-tuning experiments, whose results were consistent with those of the information flow experiments. The convergence of findings from both experiments corroborates the existence of MRE. Furthermore, we extended the application of MRE to prompt learning, utilizing word-level information as a verbalizer to bolster the model's prediction of text-level classification labels. In our final experiment, the F1-score significantly surpassed the baseline in five out of six datasets, further validating the notion that word-level information enhances the language model's comprehension of the text as a whole.
Authors: Osman Alperen Kora\c{s}, J\"org Schl\"otterer, Christin Seifert
Abstract: We present a detailed replication study of the BASS framework, an abstractive summarization system based on the notion of Unified Semantic Graphs. Our investigation includes challenges in replicating key components and an ablation study to systematically isolate error sources rooted in replicating novel components. Our findings reveal discrepancies in performance compared to the original work. We highlight the significance of paying careful attention even to reasonably omitted details for replicating advanced frameworks like BASS, and emphasize key practices for writing replicable papers.
Authors: Miaomiao Li, Jiaqi Zhu, Yang Wang, Yi Yang, Yilin Li, Hongan Wang
Abstract: Weakly supervised text classification (WSTC), also called zero-shot or dataless text classification, has attracted increasing attention due to its applicability in classifying a mass of texts within the dynamic and open Web environment, since it requires only a limited set of seed words (label names) for each category instead of labeled data. With the help of recently popular prompting Pre-trained Language Models (PLMs), many studies leveraged manually crafted and/or automatically identified verbalizers to estimate the likelihood of categories, but they failed to differentiate the effects of these category-indicative words, let alone capture their correlations and realize adaptive adjustments according to the unlabeled corpus. In this paper, in order to let the PLM effectively understand each category, we at first propose a novel form of rule-based knowledge using logical expressions to characterize the meanings of categories. Then, we develop a prompting PLM-based approach named RulePrompt for the WSTC task, consisting of a rule mining module and a rule-enhanced pseudo label generation module, plus a self-supervised fine-tuning module to make the PLM align with this task. Within this framework, the inaccurate pseudo labels assigned to texts and the imprecise logical rules associated with categories mutually enhance each other in an alternative manner. That establishes a self-iterative closed loop of knowledge (rule) acquisition and utilization, with seed words serving as the starting point. Extensive experiments validate the effectiveness and robustness of our approach, which markedly outperforms state-of-the-art weakly supervised methods. What is more, our approach yields interpretable category rules, proving its advantage in disambiguating easily-confused categories.
Authors: Kazuki Kawamura, Jun Rekimoto
Abstract: Since humans can listen to audio and watch videos at faster speeds than actually observed, we often listen to or watch these pieces of content at higher playback speeds to increase the time efficiency of content comprehension. To further utilize this capability, systems that automatically adjust the playback speed according to the user's condition and the type of content to assist in more efficient comprehension of time-series content have been developed. However, there is still room for these systems to further extend human speed-listening ability by generating speech with playback speed optimized for even finer time units and providing it to humans. In this study, we determine whether humans can hear the optimized speech and propose a system that automatically adjusts playback speed at units as small as phonemes while ensuring speech intelligibility. The system uses the speech recognizer score as a proxy for how well a human can hear a certain unit of speech and maximizes the speech playback speed to the extent that a human can hear. This method can be used to produce fast but intelligible speech. In the evaluation experiment, we compared the speech played back at a constant fast speed and the flexibly speed-up speech generated by the proposed method in a blind test and confirmed that the proposed method produced speech that was easier to listen to.
Authors: Bin Zhang, Yuxiao Ye, Guoqing Du, Xiaoru Hu, Zhishuai Li, Sun Yang, Chi Harold Liu, Rui Zhao, Ziyue Li, Hangyu Mao
Abstract: Large Language Models (LLMs) have emerged as a powerful tool in advancing the Text-to-SQL task, significantly outperforming traditional methods. Nevertheless, as a nascent research field, there is still no consensus on the optimal prompt templates and design frameworks. Additionally, existing benchmarks inadequately explore the performance of LLMs across the various sub-tasks of the Text-to-SQL process, which hinders the assessment of LLMs' cognitive capabilities and the optimization of LLM-based solutions.To address the aforementioned issues, we firstly construct a new dataset designed to mitigate the risk of overfitting in LLMs. Then we formulate five evaluation tasks to comprehensively assess the performance of diverse methods across various LLMs throughout the Text-to-SQL process.Our study highlights the performance disparities among LLMs and proposes optimal in-context learning solutions tailored to each task. These findings offer valuable insights for enhancing the development of LLM-based Text-to-SQL systems.
Authors: Zhitao He, Pengfei Cao, Chenhao Wang, Zhuoran Jin, Yubo Chen, Jiexin Xu, Huaijun Li, Xiaojian Jiang, Kang Liu, Jun Zhao
Abstract: With the development of deep learning, natural language processing technology has effectively improved the efficiency of various aspects of the traditional judicial industry. However, most current efforts focus solely on individual judicial stage, overlooking cross-stage collaboration. As the autonomous agents powered by large language models are becoming increasingly smart and able to make complex decisions in real-world settings, offering new insights for judicial intelligence. In this paper, (1) we introduce SimuCourt, a judicial benchmark that encompasses 420 judgment documents from real-world, spanning the three most common types of judicial cases, and a novel task Judicial Decision-Making to evaluate the judicial analysis and decision-making power of agents. To support this task, we construct a large-scale judicial knowledge base, JudicialKB, with multiple legal knowledge. (2) we propose a novel multi-agent framework, AgentsCourt. Our framework follows the real-world classic court trial process, consisting of court debate simulation, legal information retrieval and judgement refinement to simulate the decision-making of judge. (3) we perform extensive experiments, the results demonstrate that, our framework outperforms the existing advanced methods in various aspects, especially in generating legal grounds, where our model achieves significant improvements of 8.6% and 9.1% F1 score in the first and second instance settings, respectively.
Authors: Sungho Ko, Hyunjin Cho, Hyungjoo Chae, Jinyoung Yeo, Dongha Lee
Abstract: Recent studies have investigated utilizing Knowledge Graphs (KGs) to enhance Quesetion Answering (QA) performance of Large Language Models (LLMs), yet structured KG verbalization remains challengin. Existing methods, such as triple-form or free-form textual conversion of triple-form facts, encounter several issues. These include reduced evidence density due to duplicated entities or relationships, and reduced evidence clarity due to an inability to emphasize crucial evidence. To address these issues, we propose EFSum, an Evidence-focused Fact Summarization framework for enhanced QA with knowledge-augmented LLMs. We optimize an open-source LLM as a fact summarizer through distillation and preference alignment. Our extensive experiments show that EFSum improves LLM's zero-shot QA performance, and it is possible to ensure both the helpfulness and faithfulness of the summary.
Authors: Dong Yao, Asaad Alghamdi, Qingrong Xia, Xiaoye Qu, Xinyu Duan, Zhefeng Wang, Yi Zheng, Baoxing Huai, Peilun Cheng, Zhou Zhao
Abstract: Sentence semantic matching is a research hotspot in natural language processing, which is considerably significant in various key scenarios, such as community question answering, searching, chatbot, and recommendation. Since most of the advanced models directly model the semantic relevance among words between two sentences while neglecting the \textit{keywords} and \textit{intents} concepts of them, DC-Match is proposed to disentangle keywords from intents and utilizes them to optimize the matching performance. Although DC-Match is a simple yet effective method for semantic matching, it highly depends on the external NER techniques to identify the keywords of sentences, which limits the performance of semantic matching for minor languages since satisfactory NER tools are usually hard to obtain. In this paper, we propose to generally and flexibly resolve the text into multi concepts for multilingual semantic matching to liberate the model from the reliance on NER models. To this end, we devise a \underline{M}ulti-\underline{C}oncept \underline{P}arsed \underline{S}emantic \underline{M}atching framework based on the pre-trained language models, abbreviated as \textbf{MCP-SM}, to extract various concepts and infuse them into the classification tokens. We conduct comprehensive experiments on English datasets QQP and MRPC, and Chinese dataset Medical-SM. Besides, we experiment on Arabic datasets MQ2Q and XNLI, the outstanding performance further prove MCP-SM's applicability in low-resource languages.
Authors: Bosheng Ding, Chengwei Qin, Ruochen Zhao, Tianze Luo, Xinze Li, Guizhen Chen, Wenhan Xia, Junjie Hu, Anh Tuan Luu, Shafiq Joty
Abstract: In the rapidly evolving field of machine learning (ML), data augmentation (DA) has emerged as a pivotal technique for enhancing model performance by diversifying training examples without the need for additional data collection. This survey explores the transformative impact of Large Language Models (LLMs) on DA, particularly addressing the unique challenges and opportunities they present in the context of natural language processing (NLP) and beyond. From a data perspective and a learning perspective, we examine various strategies that utilize Large Language Models for data augmentation, including a novel exploration of learning paradigms where LLM-generated data is used for further training. Additionally, this paper delineates the primary challenges faced in this domain, ranging from controllable data augmentation to multi modal data augmentation. This survey highlights the paradigm shift introduced by LLMs in DA, aims to serve as a foundational guide for researchers and practitioners in this field.
Authors: Vipin Vijayan, Braeden Bowen, Scott Grigsby, Timothy Anderson, Jeremy Gwinnup
Abstract: A good evaluation framework should evaluate multimodal machine translation (MMT) models by measuring 1) their use of visual information to aid in the translation task and 2) their ability to translate complex sentences such as done for text-only machine translation. However, most current work in MMT is evaluated against the Multi30k testing sets, which do not measure these properties. Namely, the use of visual information by the MMT model cannot be shown directly from the Multi30k test set results and the sentences in Multi30k are are image captions, i.e., short, descriptive sentences, as opposed to complex sentences that typical text-only machine translation models are evaluated against. Therefore, we propose that MMT models be evaluated using 1) the CoMMuTE evaluation framework, which measures the use of visual information by MMT models, 2) the text-only WMT news translation task test sets, which evaluates translation performance against complex sentences, and 3) the Multi30k test sets, for measuring MMT model performance against a real MMT dataset. Finally, we evaluate recent MMT models trained solely against the Multi30k dataset against our proposed evaluation framework and demonstrate the dramatic drop performance against text-only testing sets compared to recent text-only MT models.
Authors: Anmol Goel, Nico Daheim, Iryna Gurevych
Abstract: Reframing a negative into a positive thought is at the crux of several cognitive approaches to mental health and psychotherapy that could be made more accessible by large language model-based solutions. Such reframing is typically non-trivial and requires multiple rationalization steps to uncover the underlying issue of a negative thought and transform it to be more positive. However, this rationalization process is currently neglected by both datasets and models which reframe thoughts in one step. In this work, we address this gap by augmenting open-source datasets for positive text rewriting with synthetically-generated Socratic rationales using a novel framework called \textsc{SocraticReframe}. \textsc{SocraticReframe} uses a sequence of question-answer pairs to rationalize the thought rewriting process. We show that such Socratic rationales significantly improve positive text rewriting for different open-source LLMs according to both automatic and human evaluations guided by criteria from psychotherapy research.
Authors: Zhengliang Shi, Shen Gao, Xiuyi Chen, Lingyong Yan, Haibo Shi, Dawei Yin, Zhumin Chen, Pengjie Ren, Suzan Verberne, Zhaochun Ren
Abstract: Tool learning empowers large language models (LLMs) as agents to use external tools to extend their capability. Existing methods employ one single LLM-based agent to iteratively select and execute tools, thereafter incorporating the result into the next action prediction. However, they still suffer from potential performance degradation when addressing complex tasks due to: (1) the limitation of the inherent capability of a single LLM to perform diverse actions, and (2) the struggle to adaptively correct mistakes when the task fails. To mitigate these problems, we propose the ConAgents, a Cooperative and interactive Agents framework, which modularizes the workflow of tool learning into Grounding, Execution, and Observing agents. We also introduce an iterative calibration (IterCali) method, enabling the agents to adapt themselves based on the feedback from the tool environment. Experiments conducted on three datasets demonstrate the superiority of our ConAgents (e.g., 6 point improvement over the SOTA baseline). We further provide fine-granularity analysis for the efficiency and consistency of our framework.
Authors: Vipin Vijayan, Braeden Bowen, Scott Grigsby, Timothy Anderson, Jeremy Gwinnup
Abstract: While most current work in multimodal machine translation (MMT) uses the Multi30k dataset for training and evaluation, we find that the resulting models overfit to the Multi30k dataset to an extreme degree. Consequently, these models perform very badly when evaluated against typical text-only testing sets such as the WMT newstest datasets. In order to perform well on both Multi30k and typical text-only datasets, we use a performant text-only machine translation (MT) model as the starting point of our MMT model. We add vision-text adapter layers connected via gating mechanisms to the MT model, and incrementally transform the MT model into an MMT model by 1) pre-training using vision-based masking of the source text and 2) fine-tuning on Multi30k.
Authors: Braeden Bowen, Vipin Vijayan, Scott Grigsby, Timothy Anderson, Jeremy Gwinnup
Abstract: The challenge of visual grounding and masking in multimodal machine translation (MMT) systems has encouraged varying approaches to the detection and selection of visually-grounded text tokens for masking. We introduce new methods for detection of visually and contextually relevant (concrete) tokens from source sentences, including detection with natural language processing (NLP), detection with object detection, and a joint detection-verification technique. We also introduce new methods for selection of detected tokens, including shortest $n$ tokens, longest $n$ tokens, and all detected concrete tokens. We utilize the GRAM MMT architecture to train models against synthetically collated multimodal datasets of source images with masked sentences, showing performance improvements and improved usage of visual context during translation tasks over the baseline model.
Authors: Yuqi Zhu, Shuofei Qiao, Yixin Ou, Shumin Deng, Ningyu Zhang, Shiwei Lyu, Yue Shen, Lei Liang, Jinjie Gu, Huajun Chen
Abstract: Large Language Models (LLMs) have demonstrated great potential in complex reasoning tasks, yet they fall short when tackling more sophisticated challenges, especially when interacting with environments through generating executable actions. This inadequacy primarily stems from the lack of built-in action knowledge in language agents, which fails to effectively guide the planning trajectories during task solving and results in planning hallucination. To address this issue, we introduce KnowAgent, a novel approach designed to enhance the planning capabilities of LLMs by incorporating explicit action knowledge. Specifically, KnowAgent employs an action knowledge base and a knowledgeable self-learning strategy to constrain the action path during planning, enabling more reasonable trajectory synthesis, and thereby enhancing the planning performance of language agents. Experimental results on HotpotQA and ALFWorld based on various backbone models demonstrate that KnowAgent can achieve comparable or superior performance to existing baselines. Further analysis indicates the effectiveness of KnowAgent in terms of planning hallucinations mitigation. Code is available in https://github.com/zjunlp/KnowAgent.
Authors: Chuanqi Cheng, Quan Tu, Wei Wu, Shuo Shang, Cunli Mao, Zhengtao Yu, Rui Yan
Abstract: Personalized dialogue systems have gained significant attention in recent years for their ability to generate responses in alignment with different personas. However, most existing approaches rely on pre-defined personal profiles, which are not only time-consuming and labor-intensive to create but also lack flexibility. We propose In-Dialogue Learning (IDL), a fine-tuning framework that enhances the ability of pre-trained large language models to leverage dialogue history to characterize persona for completing personalized dialogue generation tasks without pre-defined profiles. Our experiments on three datasets demonstrate that IDL brings substantial improvements, with BLEU and ROUGE scores increasing by up to 200% and 247%, respectively. Additionally, the results of human evaluations further validate the efficacy of our proposed method.
Authors: Flor Miriam Plaza-del-Arco, Amanda Cercas Curry, Alba Curry, Gavin Abercrombie, Dirk Hovy
Abstract: Large language models (LLMs) reflect societal norms and biases, especially about gender. While societal biases and stereotypes have been extensively researched in various NLP applications, there is a surprising gap for emotion analysis. However, emotion and gender are closely linked in societal discourse. E.g., women are often thought of as more empathetic, while men's anger is more socially accepted. To fill this gap, we present the first comprehensive study of gendered emotion attribution in five state-of-the-art LLMs (open- and closed-source). We investigate whether emotions are gendered, and whether these variations are based on societal stereotypes. We prompt the models to adopt a gendered persona and attribute emotions to an event like 'When I had a serious argument with a dear person'. We then analyze the emotions generated by the models in relation to the gender-event pairs. We find that all models consistently exhibit gendered emotions, influenced by gender stereotypes. These findings are in line with established research in psychology and gender studies. Our study sheds light on the complex societal interplay between language, gender, and emotion. The reproduction of emotion stereotypes in LLMs allows us to use those models to study the topic in detail, but raises questions about the predictive use of those same LLMs for emotion applications.
Authors: Kaiyan Zhang, Jianyu Wang, Ermo Hua, Biqing Qi, Ning Ding, Bowen Zhou
Abstract: With the advancement of language models (LMs), their exposure to private data is increasingly inevitable, and their deployment (especially for smaller ones) on personal devices, such as PCs and smartphones, has become a prevailing trend. In contexts laden with user information, enabling models to both safeguard user privacy and execute commands efficiently emerges as an essential research imperative. In this paper, we propose CoGenesis, a collaborative generation framework integrating large (hosted on cloud infrastructure) and small models (deployed on local devices) to address privacy concerns logically. Initially, we design a pipeline to create personalized writing instruction datasets enriched with extensive context details as the testbed of this research issue. Subsequently, we introduce two variants of CoGenesis based on sketch and logits respectively. Our experimental findings, based on our synthesized dataset and two additional open-source datasets, indicate that: 1) Large-scale models perform well when provided with user context but struggle in the absence of such context. 2) While specialized smaller models fine-tuned on the synthetic dataset show promise, they still lag behind their larger counterparts. 3) Our CoGenesis framework, utilizing mixed-scale models, showcases competitive performance, providing a feasible solution to privacy issues.
Authors: Hitesh Golchha, Sahil Yerawar, Dhruvesh Patel, Soham Dan, Keerthiram Murugesan
Abstract: Real-world sequential decision making is characterized by sparse rewards and large decision spaces, posing significant difficulty for experiential learning systems like $\textit{tabula rasa}$ reinforcement learning (RL) agents. Large Language Models (LLMs), with a wealth of world knowledge, can help RL agents learn quickly and adapt to distribution shifts. In this work, we introduce Language Guided Exploration (LGE) framework, which uses a pre-trained language model (called GUIDE ) to provide decision-level guidance to an RL agent (called EXPLORER). We observe that on ScienceWorld (Wang et al.,2022), a challenging text environment, LGE outperforms vanilla RL agents significantly and also outperforms other sophisticated methods like Behaviour Cloning and Text Decision Transformer.
Authors: Chenglei Si, Yanzhe Zhang, Zhengyuan Yang, Ruibo Liu, Diyi Yang
Abstract: Generative AI has made rapid advancements in recent years, achieving unprecedented capabilities in multimodal understanding and code generation. This can enable a new paradigm of front-end development, in which multimodal LLMs might directly convert visual designs into code implementations. In this work, we formalize this as a Design2Code task and conduct comprehensive benchmarking. Specifically, we manually curate a benchmark of 484 diverse real-world webpages as test cases and develop a set of automatic evaluation metrics to assess how well current multimodal LLMs can generate the code implementations that directly render into the given reference webpages, given the screenshots as input. We also complement automatic metrics with comprehensive human evaluations. We develop a suite of multimodal prompting methods and show their effectiveness on GPT-4V and Gemini Pro Vision. We further finetune an open-source Design2Code-18B model that successfully matches the performance of Gemini Pro Vision. Both human evaluation and automatic metrics show that GPT-4V performs the best on this task compared to other models. Moreover, annotators think GPT-4V generated webpages can replace the original reference webpages in 49% of cases in terms of visual appearance and content; and perhaps surprisingly, in 64% of cases GPT-4V generated webpages are considered better than the original reference webpages. Our fine-grained break-down metrics indicate that open-source models mostly lag in recalling visual elements from the input webpages and in generating correct layout designs, while aspects like text content and coloring can be drastically improved with proper finetuning.
Authors: Arda Uzuno\u{g}lu, Abdalfatah Rashid Safa, G\"ozde G\"ul \c{S}ahin
Abstract: Recently, there has been growing interest within the community regarding whether large language models are capable of planning or executing plans. However, most prior studies use LLMs to generate high-level plans for simplified scenarios lacking linguistic complexity and domain diversity, limiting analysis of their planning abilities. These setups constrain evaluation methods (e.g., predefined action space), architectural choices (e.g., only generative models), and overlook the linguistic nuances essential for realistic analysis. To tackle this, we present PARADISE, an abductive reasoning task using Q\&A format on practical procedural text sourced from wikiHow. It involves warning and tip inference tasks directly associated with goals, excluding intermediary steps, with the aim of testing the ability of the models to infer implicit knowledge of the plan solely from the given goal. Our experiments, utilizing fine-tuned language models and zero-shot prompting, reveal the effectiveness of task-specific small models over large language models in most scenarios. Despite advancements, all models fall short of human performance. Notably, our analysis uncovers intriguing insights, such as variations in model behavior with dropped keywords, struggles of BERT-family and GPT-4 with physical and abstract goals, and the proposed tasks offering valuable prior knowledge for other unseen procedural tasks. The PARADISE dataset and associated resources are publicly available for further research exploration with https://github.com/GGLAB-KU/paradise.
Authors: Akari Asai, Zexuan Zhong, Danqi Chen, Pang Wei Koh, Luke Zettlemoyer, Hannaneh Hajishirzi, Wen-tau Yih
Abstract: Parametric language models (LMs), which are trained on vast amounts of web data, exhibit remarkable flexibility and capability. However, they still face practical challenges such as hallucinations, difficulty in adapting to new data distributions, and a lack of verifiability. In this position paper, we advocate for retrieval-augmented LMs to replace parametric LMs as the next generation of LMs. By incorporating large-scale datastores during inference, retrieval-augmented LMs can be more reliable, adaptable, and attributable. Despite their potential, retrieval-augmented LMs have yet to be widely adopted due to several obstacles: specifically, current retrieval-augmented LMs struggle to leverage helpful text beyond knowledge-intensive tasks such as question answering, have limited interaction between retrieval and LM components, and lack the infrastructure for scaling. To address these, we propose a roadmap for developing general-purpose retrieval-augmented LMs. This involves a reconsideration of datastores and retrievers, the exploration of pipelines with improved retriever-LM interaction, and significant investment in infrastructure for efficient training and inference.
Authors: Hossein Aboutalebi, Hwanjun Song, Yusheng Xie, Arshit Gupta, Justin Sun, Hang Su, Igor Shalyminov, Nikolaos Pappas, Siffi Singh, Saab Mansour
Abstract: Development of multimodal interactive systems is hindered by the lack of rich, multimodal (text, images) conversational data, which is needed in large quantities for LLMs. Previous approaches augment textual dialogues with retrieved images, posing privacy, diversity, and quality constraints. In this work, we introduce \textbf{M}ultimodal \textbf{A}ugmented \textbf{G}enerative \textbf{I}mages \textbf{D}ialogues (MAGID), a framework to augment text-only dialogues with diverse and high-quality images. Subsequently, a diffusion model is applied to craft corresponding images, ensuring alignment with the identified text. Finally, MAGID incorporates an innovative feedback loop between an image description generation module (textual LLM) and image quality modules (addressing aesthetics, image-text matching, and safety), that work in tandem to generate high-quality and multi-modal dialogues. We compare MAGID to other SOTA baselines on three dialogue datasets, using automated and human evaluation. Our results show that MAGID is comparable to or better than baselines, with significant improvements in human evaluation, especially against retrieval baselines where the image database is small.
Authors: Qian Wan, Xin Feng, Yining Bei, Zhiqi Gao, Zhicong Lu
Abstract: Human emotions are essentially molded by lived experiences, from which we construct personalised meaning. The engagement in such meaning-making process has been practiced as an intervention in various psychotherapies to promote wellness. Nevertheless, to support recollecting and recounting lived experiences in everyday life remains under explored in HCI. It also remains unknown how technologies such as generative AI models can facilitate the meaning making process, and ultimately support affective mindfulness. In this paper we present Metamorpheus, an affective interface that engages users in a creative visual storytelling of emotional experiences during dreams. Metamorpheus arranges the storyline based on a dream's emotional arc, and provokes self-reflection through the creation of metaphorical images and text depictions. The system provides metaphor suggestions, and generates visual metaphors and text depictions using generative AI models, while users can apply generations to recolour and re-arrange the interface to be visually affective. Our experience-centred evaluation manifests that, by interacting with Metamorpheus, users can recall their dreams in vivid detail, through which they relive and reflect upon their experiences in a meaningful way.
Authors: Jana\'ina Mendes-Laureano, Jorge A. G\'omez-Garc\'ia, Alejandro Guerrero-L\'opez, Elisa Luque-Buzo, Juli\'an D. Arias-Londo\~no, Francisco J. Grandas-P\'erez, Juan I. Godino-Llorente
Abstract: The advancement of Parkinson's Disease (PD) diagnosis through speech analysis is hindered by a notable lack of publicly available, diverse language datasets, limiting the reproducibility and further exploration of existing research. In response to this gap, we introduce a comprehensive corpus from 108 native Castilian Spanish speakers, comprising 55 healthy controls and 53 individuals diagnosed with PD, all of whom were under pharmacological treatment and recorded in their medication-optimized state. This unique dataset features a wide array of speech tasks, including sustained phonation of the five Spanish vowels, diadochokinetic tests, 16 listen-and-repeat utterances, and free monologues. The dataset emphasizes accuracy and reliability through specialist manual transcriptions of the listen-and-repeat tasks and utilizes Whisper for automated monologue transcriptions, making it the most complete public corpus of Parkinsonian speech, and the first in Castillian Spanish. NeuroVoz is composed by 2,903 audio recordings averaging $26.88 \pm 3.35$ recordings per participant, offering a substantial resource for the scientific exploration of PD's impact on speech. This dataset has already underpinned several studies, achieving a benchmark accuracy of 89% in PD speech pattern identification, indicating marked speech alterations attributable to PD. Despite these advances, the broader challenge of conducting a language-agnostic, cross-corpora analysis of Parkinsonian speech patterns remains an open area for future research. This contribution not only fills a critical void in PD speech analysis resources but also sets a new standard for the global research community in leveraging speech as a diagnostic tool for neurodegenerative diseases.
Authors: Lingjiao Chen, Jared Quincy Davis, Boris Hanin, Peter Bailis, Ion Stoica, Matei Zaharia, James Zou
Abstract: Many recent state-of-the-art results in language tasks were achieved using compound systems that perform multiple Large Language Model (LLM) calls and aggregate their responses. However, there is little understanding of how the number of LLM calls -- e.g., when asking the LLM to answer each question multiple times and taking a consensus -- affects such a compound system's performance. In this paper, we initiate the study of scaling laws of compound inference systems. We analyze, theoretically and empirically, how the number of LLM calls affects the performance of one-layer Voting Inference Systems -- one of the simplest compound systems, which aggregates LLM responses via majority voting. We find empirically that across multiple language tasks, surprisingly, Voting Inference Systems' performance first increases but then decreases as a function of the number of LLM calls. Our theoretical results suggest that this non-monotonicity is due to the diversity of query difficulties within a task: more LLM calls lead to higher performance on "easy" queries, but lower performance on "hard" queries, and non-monotone behavior emerges when a task contains both types of queries. This insight then allows us to compute, from a small number of samples, the number of LLM calls that maximizes system performance, and define a scaling law of Voting Inference Systems. Experiments show that our scaling law can predict the performance of Voting Inference Systems and find the optimal number of LLM calls to make.
Authors: Zixuan Liu, Xiaolin Sun, Zizhan Zheng
Abstract: The rapidly increasing capabilities of large language models (LLMs) raise an urgent need to align AI systems with diverse human preferences to simultaneously enhance their usefulness and safety, despite the often conflicting nature of these goals. To address this important problem, a promising approach is to enforce a safety constraint at the fine-tuning stage through a constrained Reinforcement Learning from Human Feedback (RLHF) framework. This approach, however, is computationally expensive and often unstable. In this work, we introduce Constrained DPO (C-DPO), a novel extension of the recently proposed Direct Preference Optimization (DPO) approach for fine-tuning LLMs that is both efficient and lightweight. By integrating dual gradient descent and DPO, our method identifies a nearly optimal trade-off between helpfulness and harmlessness without using reinforcement learning. Empirically, our approach provides a safety guarantee to LLMs that is missing in DPO while achieving significantly higher rewards under the same safety constraint compared to a recently proposed safe RLHF approach. Warning: This paper contains example data that may be offensive or harmful.
Authors: Aashaka Desai, Maartje De Meulder, Julie A. Hochgesang, Annemarie Kocab, Alex X. Lu
Abstract: Growing research in sign language recognition, generation, and translation AI has been accompanied by calls for ethical development of such technologies. While these works are crucial to helping individual researchers do better, there is a notable lack of discussion of systemic biases or analysis of rhetoric that shape the research questions and methods in the field, especially as it remains dominated by hearing non-signing researchers. Therefore, we conduct a systematic review of 101 recent papers in sign language AI. Our analysis identifies significant biases in the current state of sign language AI research, including an overfocus on addressing perceived communication barriers, a lack of use of representative datasets, use of annotations lacking linguistic foundations, and development of methods that build on flawed models. We take the position that the field lacks meaningful input from Deaf stakeholders, and is instead driven by what decisions are the most convenient or perceived as important to hearing researchers. We end with a call to action: the field must make space for Deaf researchers to lead the conversation in sign language AI.
Authors: Yutong Li, Lu Chen, Aiwei Liu, Kai Yu, Lijie Wen
Abstract: The literature review is an indispensable step in the research process. It provides the benefit of comprehending the research problem and understanding the current research situation while conducting a comparative analysis of prior works. However, literature summary is challenging and time consuming. The previous LLM-based studies on literature review mainly focused on the complete process, including literature retrieval, screening, and summarization. However, for the summarization step, simple CoT method often lacks the ability to provide extensive comparative summary. In this work, we firstly focus on the independent literature summarization step and introduce ChatCite, an LLM agent with human workflow guidance for comparative literature summary. This agent, by mimicking the human workflow, first extracts key elements from relevant literature and then generates summaries using a Reflective Incremental Mechanism. In order to better evaluate the quality of the generated summaries, we devised a LLM-based automatic evaluation metric, G-Score, in refer to the human evaluation criteria. The ChatCite agent outperformed other models in various dimensions in the experiments. The literature summaries generated by ChatCite can also be directly used for drafting literature reviews.
Authors: Weizhi Wang, Khalil Mrini, Linjie Yang, Sateesh Kumar, Yu Tian, Xifeng Yan, Heng Wang
Abstract: We propose a novel framework for filtering image-text data by leveraging fine-tuned Multimodal Language Models (MLMs). Our approach outperforms predominant filtering methods (e.g., CLIPScore) via integrating the recent advances in MLMs. We design four distinct yet complementary metrics to holistically measure the quality of image-text data. A new pipeline is established to construct high-quality instruction data for fine-tuning MLMs as data filters. Comparing with CLIPScore, our MLM filters produce more precise and comprehensive scores that directly improve the quality of filtered data and boost the performance of pre-trained models. We achieve significant improvements over CLIPScore on popular foundation models (i.e., CLIP and BLIP2) and various downstream tasks. Our MLM filter can generalize to different models and tasks, and be used as a drop-in replacement for CLIPScore. An additional ablation study is provided to verify our design choices for the MLM filter.
Authors: Waris GillVirginia Tech, USA, Mohamed ElidrisiCisco, USA, Pallavi KalapatapuCisco, USA, Ali AnwarUniversity of Minnesota, Minneapolis, USA, Muhammad Ali GulzarVirginia Tech, USA
Abstract: Large Language Models (LLMs) like ChatGPT, Google Bard, Claude, and Llama 2 have revolutionized natural language processing and search engine dynamics. However, these models incur exceptionally high computational costs. For instance, GPT-3 consists of 175 billion parameters and inference on these models also demands billions of floating-point operations. Caching is a natural solution to reduce LLM inference costs on repeated queries. However, existing caching methods are incapable of finding semantic similarities among LLM queries, leading to unacceptable false hit-and-miss rates. This paper introduces MeanCache, a semantic cache for LLMs that identifies semantically similar queries to determine cache hit or miss. Using MeanCache, the response to a user's semantically similar query can be retrieved from a local cache rather than re-querying the LLM, thus reducing costs, service provider load, and environmental impact. MeanCache leverages Federated Learning (FL) to collaboratively train a query similarity model in a distributed manner across numerous users without violating privacy. By placing a local cache in each user's device and using FL, MeanCache reduces the latency and costs and enhances model performance, resulting in lower cache false hit rates. Our experiments, benchmarked against the GPTCache, reveal that MeanCache attains an approximately 17% higher F-score and a 20% increase in precision during semantic cache hit-and-miss decisions. Furthermore, MeanCache reduces the storage requirement by 83% and accelerates semantic cache hit-and-miss decisions by 11%, while still surpassing GPTCache.
Authors: Son The Nguyen, Niranjan Uma Naresh, Theja Tulabandhula
Abstract: This paper addresses the challenges of aligning large language models (LLMs) with human values via preference learning (PL), with a focus on the issues of incomplete and corrupted data in preference datasets. We propose a novel method for robustly and completely recalibrating values within these datasets to enhance LLMs resilience against the issues. In particular, we devise a guaranteed polynomial time ranking algorithm that robustifies several existing models, such as the classic Bradley--Terry--Luce (BTL) (Bradley and Terry, 1952) model and certain generalizations of it. To the best of our knowledge, our present work is the first to propose an algorithm that provably recovers an {\epsilon}-optimal ranking with high probability while allowing as large as O(n) perturbed pairwise comparison results per model response. Furthermore, we show robust recovery results in the partially observed setting. Our experiments confirm that our algorithms handle adversarial noise and unobserved comparisons well in both general and LLM preference dataset settings. This work contributes to the development and scaling of more reliable and ethically aligned AI models by equipping the dataset curation pipeline with the ability to handle missing and maliciously manipulated inputs.
Authors: Joy He-Yueya, Noah D. Goodman, Emma Brunskill
Abstract: Creating effective educational materials generally requires expensive and time-consuming studies of student learning outcomes. To overcome this barrier, one idea is to build computational models of student learning and use them to optimize instructional materials. However, it is difficult to model the cognitive processes of learning dynamics. We propose an alternative approach that uses Language Models (LMs) as educational experts to assess the impact of various instructions on learning outcomes. Specifically, we use GPT-3.5 to evaluate the overall effect of instructional materials on different student groups and find that it can replicate well-established educational findings such as the Expertise Reversal Effect and the Variability Effect. This demonstrates the potential of LMs as reliable evaluators of educational content. Building on this insight, we introduce an instruction optimization approach in which one LM generates instructional materials using the judgments of another LM as a reward function. We apply this approach to create math word problem worksheets aimed at maximizing student learning gains. Human teachers' evaluations of these LM-generated worksheets show a significant alignment between the LM judgments and human teacher preferences. We conclude by discussing potential divergences between human and LM opinions and the resulting pitfalls of automating instructional design.
Authors: Philipp J. R\"osch, Norbert Oswald, Michaela Geierhos, Jind\v{r}ich Libovick\'y
Abstract: Current multimodal models leveraging contrastive learning often face limitations in developing fine-grained conceptual understanding. This is due to random negative samples during pretraining, causing almost exclusively very dissimilar concepts to be compared in the loss function. Consequently, the models struggle with fine-grained semantic differences. To address this problem, we introduce a novel pretraining method incorporating synthetic hard negative text examples. The hard negatives permute terms corresponding to visual concepts, leading to a more fine-grained visual and textual concept alignment. Further, we introduce InpaintCOCO, a new challenging dataset for assessing the fine-grained alignment of colors, objects, and sizes in vision-language models. We created the dataset using generative inpainting from COCO images by changing the visual concepts so that the images no longer match their original captions. Our results show significant improvements in fine-grained concept understanding across a wide range of vision-language datasets, including our InpaintCOCO dataset.
Authors: Yoonjoo Lee, Hyeonsu B. Kang, Matt Latzke, Juho Kim, Jonathan Bragg, Joseph Chee Chang, Pao Siangliulue
Abstract: With the rapid growth of scholarly archives, researchers subscribe to "paper alert" systems that periodically provide them with recommendations of recently published papers that are similar to previously collected papers. However, researchers sometimes struggle to make sense of nuanced connections between recommended papers and their own research context, as existing systems only present paper titles and abstracts. To help researchers spot these connections, we present PaperWeaver, an enriched paper alerts system that provides contextualized text descriptions of recommended papers based on user-collected papers. PaperWeaver employs a computational method based on Large Language Models (LLMs) to infer users' research interests from their collected papers, extract context-specific aspects of papers, and compare recommended and collected papers on these aspects. Our user study (N=15) showed that participants using PaperWeaver were able to better understand the relevance of recommended papers and triage them more confidently when compared to a baseline that presented the related work sections from recommended papers.
Authors: Stefan Hackmann, Haniyeh Mahmoudian, Mark Steadman, Michael Schmidt
Abstract: The emergence of large language models (LLMs) has revolutionized numerous applications across industries. However, their "black box" nature often hinders the understanding of how they make specific decisions, raising concerns about their transparency, reliability, and ethical use. This study presents a method to improve the explainability of LLMs by varying individual words in prompts to uncover their statistical impact on the model outputs. This approach, inspired by permutation importance for tabular data, masks each word in the system prompt and evaluates its effect on the outputs based on the available text scores aggregated over multiple user inputs. Unlike classical attention, word importance measures the impact of prompt words on arbitrarily-defined text scores, which enables decomposing the importance of words into the specific measures of interest--including bias, reading level, verbosity, etc. This procedure also enables measuring impact when attention weights are not available. To test the fidelity of this approach, we explore the effect of adding different suffixes to multiple different system prompts and comparing subsequent generations with different large language models. Results show that word importance scores are closely related to the expected suffix importances for multiple scoring functions.
Authors: Zeqian Ju, Yuancheng Wang, Kai Shen, Xu Tan, Detai Xin, Dongchao Yang, Yanqing Liu, Yichong Leng, Kaitao Song, Siliang Tang, Zhizheng Wu, Tao Qin, Xiang-Yang Li, Wei Ye, Shikun Zhang, Jiang Bian, Lei He, Jinyu Li, Sheng Zhao
Abstract: While recent large-scale text-to-speech (TTS) models have achieved significant progress, they still fall short in speech quality, similarity, and prosody. Considering speech intricately encompasses various attributes (e.g., content, prosody, timbre, and acoustic details) that pose significant challenges for generation, a natural idea is to factorize speech into individual subspaces representing different attributes and generate them individually. Motivated by it, we propose NaturalSpeech 3, a TTS system with novel factorized diffusion models to generate natural speech in a zero-shot way. Specifically, 1) we design a neural codec with factorized vector quantization (FVQ) to disentangle speech waveform into subspaces of content, prosody, timbre, and acoustic details; 2) we propose a factorized diffusion model to generate attributes in each subspace following its corresponding prompt. With this factorization design, NaturalSpeech 3 can effectively and efficiently model the intricate speech with disentangled subspaces in a divide-and-conquer way. Experiments show that NaturalSpeech 3 outperforms the state-of-the-art TTS systems on quality, similarity, prosody, and intelligibility. Furthermore, we achieve better performance by scaling to 1B parameters and 200K hours of training data.
Authors: Peng Qi, Zehong Yan, Wynne Hsu, Mong Li Lee
Abstract: Misinformation is a prevalent societal issue due to its potential high risks. Out-of-context (OOC) misinformation, where authentic images are repurposed with false text, is one of the easiest and most effective ways to mislead audiences. Current methods focus on assessing image-text consistency but lack convincing explanations for their judgments, which is essential for debunking misinformation. While Multimodal Large Language Models (MLLMs) have rich knowledge and innate capability for visual reasoning and explanation generation, they still lack sophistication in understanding and discovering the subtle crossmodal differences. In this paper, we introduce SNIFFER, a novel multimodal large language model specifically engineered for OOC misinformation detection and explanation. SNIFFER employs two-stage instruction tuning on InstructBLIP. The first stage refines the model's concept alignment of generic objects with news-domain entities and the second stage leverages language-only GPT-4 generated OOC-specific instruction data to fine-tune the model's discriminatory powers. Enhanced by external tools and retrieval, SNIFFER not only detects inconsistencies between text and image but also utilizes external knowledge for contextual verification. Our experiments show that SNIFFER surpasses the original MLLM by over 40% and outperforms state-of-the-art methods in detection accuracy. SNIFFER also provides accurate and persuasive explanations as validated by quantitative and human evaluations.
Authors: Nathaniel Li, Alexander Pan, Anjali Gopal, Summer Yue, Daniel Berrios, Alice Gatti, Justin D. Li, Ann-Kathrin Dombrowski, Shashwat Goel, Long Phan, Gabriel Mukobi, Nathan Helm-Burger, Rassin Lababidi, Lennart Justen, Andrew B. Liu, Michael Chen, Isabelle Barrass, Oliver Zhang, Xiaoyuan Zhu, Rishub Tamirisa, Bhrugu Bharathi, Adam Khoja, Ariel Herbert-Voss, Cort B. Breuer, Andy Zou, Mantas Mazeika, Zifan Wang, Palash Oswal, Weiran Liu, Adam A. Hunt, Justin Tienken-Harder, Kevin Y. Shih, Kemper Talley, John Guan, Russell Kaplan, Ian Steneker, David Campbell, Brad Jokubaitis, Alex Levinson, Jean Wang, William Qian, Kallol Krishna Karmakar, Steven Basart, Stephen Fitz, Mindy Levine, Ponnurangam Kumaraguru, Uday Tupakula, Vijay Varadharajan, Yan Shoshitaishvili, Jimmy Ba, Kevin M. Esvelt, Alexandr Wang, Dan Hendrycks
Abstract: The White House Executive Order on Artificial Intelligence highlights the risks of large language models (LLMs) empowering malicious actors in developing biological, cyber, and chemical weapons. To measure these risks of malicious use, government institutions and major AI labs are developing evaluations for hazardous capabilities in LLMs. However, current evaluations are private, preventing further research into mitigating risk. Furthermore, they focus on only a few, highly specific pathways for malicious use. To fill these gaps, we publicly release the Weapons of Mass Destruction Proxy (WMDP) benchmark, a dataset of 4,157 multiple-choice questions that serve as a proxy measurement of hazardous knowledge in biosecurity, cybersecurity, and chemical security. WMDP was developed by a consortium of academics and technical consultants, and was stringently filtered to eliminate sensitive information prior to public release. WMDP serves two roles: first, as an evaluation for hazardous knowledge in LLMs, and second, as a benchmark for unlearning methods to remove such hazardous knowledge. To guide progress on unlearning, we develop CUT, a state-of-the-art unlearning method based on controlling model representations. CUT reduces model performance on WMDP while maintaining general capabilities in areas such as biology and computer science, suggesting that unlearning may be a concrete path towards reducing malicious use from LLMs. We release our benchmark and code publicly at https://wmdp.ai
URLs: https://wmdp.ai
Authors: Yifu Sun, Haoming Jiang
Abstract: Recently, with the help of deep learning models, significant advances have been made in different Natural Language Processing (NLP) tasks. Unfortunately, state-of-the-art models are vulnerable to noisy texts. We propose a new contextual text denoising algorithm based on the ready-to-use masked language model. The proposed algorithm does not require retraining of the model and can be integrated into any NLP system without additional training on paired cleaning training data. We evaluate our method under synthetic noise and natural noise and show that the proposed algorithm can use context information to correct noise text and improve the performance of noisy inputs in several downstream tasks.
Authors: Sebastian Vincent, Alice Dowek, Rowanne Sumner, Charlotte Blundell, Emily Preston, Chris Bayliss, Chris Oakley, Carolina Scarton
Abstract: Sensitising language models (LMs) to external context helps them to more effectively capture the speaking patterns of individuals with specific characteristics or in particular environments. This work investigates to what extent rich character and film annotations can be leveraged to personalise LMs in a scalable manner. We then explore the use of such models in evaluating context specificity in machine translation. We build LMs which leverage rich contextual information to reduce perplexity by up to 6.5% compared to a non-contextual model, and generalise well to a scenario with no speaker-specific data, relying on combinations of demographic characteristics expressed via metadata. Our findings are consistent across two corpora, one of which (Cornell-rich) is also a contribution of this paper. We then use our personalised LMs to measure the co-occurrence of extra-textual context and translation hypotheses in a machine translation setting. Our results suggest that the degree to which professional translations in our domain are context-specific can be preserved to a better extent by a contextual machine translation model than a non-contextual model, which is also reflected in the contextual model's superior reference-based scores.
Authors: Zhang Ze Yu, Lau Jia Jaw, Zhang Hui, Bryan Kian Hsiang Low
Abstract: Reinforcement Learning with Human Feedback (RLHF) has been demonstrated to significantly enhance the performance of large language models (LLMs) by aligning their outputs with desired human values through instruction tuning. However, RLHF is constrained by the expertise and productivity limitations of human evaluators. A response to this downside is to fall back to supervised fine-tuning (SFT) with additional carefully selected expert demonstrations. However, while this method has been proven to be effective, it invariably also leads to increased human-in-the-loop overhead. In this study, we propose another alternative approach: Reinforcement Learning with Generative Adversarial Feedback (RLGAF) to RLHF and SFT, which uses a generative adversarial training style to enable the LLMs to learn useful human expert demonstrations without being directly exposed to the training examples, thus enabling good generalization capabilities while preserving sample efficiency. Our preliminary findings indicate that RLGAF can help align LLMs outputs with competitive performance against RLHF and SFT, while not suffering from their respective inherent restrictions, suggesting promising avenues for further research on automating AI alignment.
Authors: Yaxin Fan, Feng Jiang, Peifeng Li, Haizhou Li
Abstract: Large language models, like ChatGPT, have shown remarkable capability in many downstream tasks, yet their ability to understand discourse structures of dialogues remains less explored, where it requires higher level capabilities of understanding and reasoning. In this paper, we aim to systematically inspect ChatGPT's performance in two discourse analysis tasks: topic segmentation and discourse parsing, focusing on its deep semantic understanding of linear and hierarchical discourse structures underlying dialogue. To instruct ChatGPT to complete these tasks, we initially craft a prompt template consisting of the task description, output format, and structured input. Then, we conduct experiments on four popular topic segmentation datasets and two discourse parsing datasets. The experimental results showcase that ChatGPT demonstrates proficiency in identifying topic structures in general-domain conversations yet struggles considerably in specific-domain conversations. We also found that ChatGPT hardly understands rhetorical structures that are more complex than topic structures. Our deeper investigation indicates that ChatGPT can give more reasonable topic structures than human annotations but only linearly parses the hierarchical rhetorical structures. In addition, we delve into the impact of in-context learning (e.g., chain-of-thought) on ChatGPT and conduct the ablation study on various prompt components, which can provide a research foundation for future work. The code is available at \url{https://github.com/yxfanSuda/GPTforDDA}.
Authors: Michael J. Q. Zhang, Eunsol Choi
Abstract: While large language models are able to retain vast amounts of world knowledge seen during pretraining, such knowledge is prone to going out of date and is nontrivial to update. Furthermore, these models are often used under temporal misalignment, tasked with answering questions about the present, despite having only been trained on data collected in the past. To mitigate the effects of temporal misalignment, we propose fact duration prediction: the task of predicting how long a given fact will remain true. In our experiments, we demonstrate that identifying which facts are prone to rapid change can help models avoid reciting outdated information and determine which predictions require seeking out up-to-date knowledge sources. We also show how modeling fact duration improves calibration for knowledge-intensive tasks, such as open-retrieval question answering, under temporal misalignment, by discarding volatile facts. Our data and code are released publicly at https://github.com/mikejqzhang/mitigating_misalignment.
URLs: https://github.com/mikejqzhang/mitigating_misalignment.
Authors: Yuchuan Tian, Hanting Chen, Xutao Wang, Zheyuan Bai, Qinghua Zhang, Ruifeng Li, Chao Xu, Yunhe Wang
Abstract: Recent releases of Large Language Models (LLMs), e.g. ChatGPT, are astonishing at generating human-like texts, but they may impact the authenticity of texts. Previous works proposed methods to detect these AI-generated texts, including simple ML classifiers, pretrained-model-based zero-shot methods, and finetuned language classification models. However, mainstream detectors always fail on short texts, like SMSes, Tweets, and reviews. In this paper, a Multiscale Positive-Unlabeled (MPU) training framework is proposed to address the difficulty of short-text detection without sacrificing long-texts. Firstly, we acknowledge the human-resemblance property of short machine texts, and rephrase AI text detection as a partial Positive-Unlabeled (PU) problem by regarding these short machine texts as partially ``unlabeled". Then in this PU context, we propose the length-sensitive Multiscale PU Loss, where a recurrent model in abstraction is used to estimate positive priors of scale-variant corpora. Additionally, we introduce a Text Multiscaling module to enrich training corpora. Experiments show that our MPU method augments detection performance on long AI-generated texts, and significantly improves short-text detection of language model detectors. Language Models trained with MPU could outcompete existing detectors on various short-text and long-text detection benchmarks. The codes are available at https://github.com/mindspore-lab/mindone/tree/master/examples/detect_chatgpt and https://github.com/YuchuanTian/AIGC_text_detector.
URLs: https://github.com/mindspore-lab/mindone/tree/master/examples/detect_chatgpt, https://github.com/YuchuanTian/AIGC_text_detector.
Authors: Jacob-Junqi Tian, David Emerson, Sevil Zanjani Miyandoab, Deval Pandya, Laleh Seyyed-Kalantari, Faiza Khan Khattak
Abstract: Prompting large language models has gained immense popularity in recent years due to the advantage of producing good results even without the need for labelled data. However, this requires prompt tuning to get optimal prompts that lead to better model performances. In this paper, we explore the use of soft-prompt tuning on sentiment classification task to quantify the biases of large language models (LLMs) such as Open Pre-trained Transformers (OPT) and Galactica language model. Since these models are trained on real-world data that could be prone to bias toward certain groups of populations, it is important to identify these underlying issues. Using soft-prompts to evaluate bias gives us the extra advantage of avoiding the human-bias injection that can be caused by manually designed prompts. We check the model biases on different sensitive attributes using the group fairness (bias) and find interesting bias patterns. Since LLMs have been used in the industry in various applications, it is crucial to identify the biases before deploying these models in practice. We open-source our pipeline and encourage industry researchers to adapt our work to their use cases.
Authors: Marco Guevara, Shan Chen, Spencer Thomas, Tafadzwa L. Chaunzwa, Idalid Franco, Benjamin Kann, Shalini Moningi, Jack Qian, Madeleine Goldstein, Susan Harper, Hugo JWL Aerts, Guergana K. Savova, Raymond H. Mak, Danielle S. Bitterman
Abstract: Social determinants of health (SDoH) have an important impact on patient outcomes but are incompletely collected from the electronic health records (EHR). This study researched the ability of large language models to extract SDoH from free text in EHRs, where they are most commonly documented, and explored the role of synthetic clinical text for improving the extraction of these scarcely documented, yet extremely valuable, clinical data. 800 patient notes were annotated for SDoH categories, and several transformer-based models were evaluated. The study also experimented with synthetic data generation and assessed for algorithmic bias. Our best-performing models were fine-tuned Flan-T5 XL (macro-F1 0.71) for any SDoH, and Flan-T5 XXL (macro-F1 0.70). The benefit of augmenting fine-tuning with synthetic data varied across model architecture and size, with smaller Flan-T5 models (base and large) showing the greatest improvements in performance (delta F1 +0.12 to +0.23). Model performance was similar on the in-hospital system dataset but worse on the MIMIC-III dataset. Our best-performing fine-tuned models outperformed zero- and few-shot performance of ChatGPT-family models for both tasks. These fine-tuned models were less likely than ChatGPT to change their prediction when race/ethnicity and gender descriptors were added to the text, suggesting less algorithmic bias (p<0.05). At the patient-level, our models identified 93.8% of patients with adverse SDoH, while ICD-10 codes captured 2.0%. Our method can effectively extracted SDoH information from clinic notes, performing better compare to GPT zero- and few-shot settings. These models could enhance real-world evidence on SDoH and aid in identifying patients needing social support.
Authors: Bin Wang, Zhengyuan Liu, Xin Huang, Fangkai Jiao, Yang Ding, Ai Ti Aw, Nancy F. Chen
Abstract: We present SeaEval, a benchmark for multilingual foundation models. In addition to characterizing how these models understand and reason with natural language, we also investigate how well they comprehend cultural practices, nuances, and values. Alongside standard accuracy metrics, we investigate the brittleness of foundation models in the dimensions of semantics and multilinguality. Our analyses span both open-sourced and closed models, leading to empirical results across classic NLP tasks, reasoning, and cultural comprehension. Key findings indicate (1) Most models exhibit varied behavior when given paraphrased instructions. (2) Many models still suffer from exposure bias (e.g., positional bias, majority label bias). (3) For questions rooted in factual, scientific, and commonsense knowledge, consistent responses are expected across multilingual queries that are semantically equivalent. Yet, most models surprisingly demonstrate inconsistent performance on these queries. (4) Multilingually-trained models have not attained "balanced multilingual" capabilities. Our endeavors underscore the need for more generalizable semantic representations and enhanced multilingual contextualization. SeaEval can serve as a launchpad for more thorough investigations and evaluations for multilingual and multicultural scenarios.
Authors: Yuhan Chen, Nuwa Xi, Yanrui Du, Haochun Wang, Jianyu Chen, Sendong Zhao, Bing Qin
Abstract: Molecule discovery serves as a cornerstone in numerous scientific domains, fueling the development of new materials and innovative drug designs. Recent developments of in-silico molecule discovery have highlighted the promising results of cross-modal techniques, which bridge molecular structures with their descriptive annotations. However, these cross-modal methods frequently encounter the issue of data scarcity, hampering their performance and application. In this paper, we address the low-resource challenge by utilizing artificially-real data generated by Large Language Models (LLMs). We first introduce a retrieval-based prompting strategy to construct high-quality pseudo data, then explore the optimal method to effectively leverage this pseudo data. Experiments show that using pseudo data for domain adaptation outperforms all existing methods, while also requiring a smaller model scale, reduced data size and lower training cost, highlighting its efficiency. Furthermore, our method shows a sustained improvement as the volume of pseudo data increases, revealing the great potential of pseudo data in advancing low-resource cross-modal molecule discovery. Our code and data are available at https://github.com/SCIR-HI/ArtificiallyR2R.
Authors: Lauren Hong, Ting Wang
Abstract: Parameter-efficient fine-tuning (PEFT) enables efficient adaptation of pre-trained language models (PLMs) to specific tasks. By tuning only a minimal set of (extra) parameters, PEFT achieves performance that is comparable to standard fine-tuning. However, despite its prevalent use, the security implications of PEFT remain largely unexplored. In this paper, we take the initial steps and present PETA, a novel trojan attack that compromises the weights of PLMs by accounting for downstream adaptation through bilevel optimization: the upper-level objective embeds the backdoor into a model while the lower-level objective simulates PEFT to both retain the PLM's task-specific performance and ensure that the backdoor persists after fine-tuning. With extensive evaluation across a variety of downstream tasks and trigger designs, we demonstrate PETA's effectiveness in terms of both attack success rate and clean accuracy, even when the attacker does not have full knowledge of the victim user's training process.
Authors: Yutaro Yamada, Yihan Bao, Andrew K. Lampinen, Jungo Kasai, Ilker Yildirim
Abstract: Large language models (LLMs) show remarkable capabilities across a variety of tasks. Despite the models only seeing text in training, several recent studies suggest that LLM representations implicitly capture aspects of the underlying grounded concepts. Here, we explore LLM representations of a particularly salient kind of grounded knowledge -- spatial relationships. We design natural-language navigation tasks and evaluate the ability of LLMs, in particular GPT-3.5-turbo, GPT-4, and Llama2 series models, to represent and reason about spatial structures. These tasks reveal substantial variability in LLM performance across different spatial structures, including square, hexagonal, and triangular grids, rings, and trees. In extensive error analysis, we find that LLMs' mistakes reflect both spatial and non-spatial factors. These findings suggest that LLMs appear to capture certain aspects of spatial structure implicitly, but room for improvement remains.
Authors: Yixin Wan, Fanyou Wu, Weijie Xu, Srinivasan H. Sengamedu
Abstract: In this work, we propose sequence-level certainty as a common theme over hallucination in Knowledge Grounded Dialogue Generation (KGDG). We explore the correlation between the level of hallucination and two types of sequence-level certainty: probabilistic certainty and semantic certainty. Empirical results reveal that a higher level of both types of sequence-level certainty in model responses is correlated with a lower level of hallucination. We further propose Certainty-based Response Ranking (CRR), a decoding-time hallucination mitigation method that ranks response candidates based on their sequence-level certainty and outputs the answer with the highest certainty level. Aligning with our definitions of sequence-level certainty, we design 2 types of CRR approaches: Probabilistic CRR (P-CRR) and Semantic CRR (S-CRR). P-CRR ranks individually sampled model responses using the arithmetic mean log-probability of the entire sequence. S-CRR approaches certainty estimation from meaning-space, and ranks model response candidates based on their semantic certainty level as measured by an entailment-based Agreement Score (AS). Through extensive experiments across 3 KGDG datasets, 3 decoding methods, and 4 different models, we validate the effectiveness of the CRR methods in reducing model hallucination.
Authors: Hui Wang, Anh Dang, Zihao Wu, Son Mac
Abstract: The advancements in Generative Artificial Intelligence (GenAI) technologies such as ChatGPT provide opportunities to enrich educational experiences, but also raise concerns about academic integrity if misused. This study aims to explore how universities and educators respond and adapt to the development of GenAI in their academic contexts by analyzing academic policies and guidelines established by top-ranked US universities regarding the use of ChatGPT in higher education. The data sources include academic policies, statements, guidelines as well as relevant resources provided by the top 100 universities in the US. Results show that the majority of these universities adopt an open but cautious approach towards the integration of GenAI. Primary concerns lie in ethical usage, accuracy, and data privacy. Most universities actively respond and provide diverse types of resources, such as syllabus templates/samples, workshops, shared articles, and one-on-one consultations, with topics focusing on general technical introduction, ethical concerns, pedagogical applications, preventive strategies, data privacy, limitations, and detective tools. The findings provide two suggestions for educators in policy-making: establish discipline-specific policies, and manage sensitive information carefully, as well as four implications for educators in teaching practices: accept its presence, align its use with learning objectives, evolve curriculum to prevent misuse, and adopt multifaceted evaluation strategies rather than relying on AI detectors.
Authors: Yuxin Zi, Hariram Veeramani, Kaushik Roy, Amit Sheth
Abstract: Natural language understanding (NLU) using neural network pipelines often requires additional context that is not solely present in the input data. Through Prior research, it has been evident that NLU benchmarks are susceptible to manipulation by neural models, wherein these models exploit statistical artifacts within the encoded external knowledge to artificially inflate performance metrics for downstream tasks. Our proposed approach, known as the Recap, Deliberate, and Respond (RDR) paradigm, addresses this issue by incorporating three distinct objectives within the neural network pipeline. Firstly, the Recap objective involves paraphrasing the input text using a paraphrasing model in order to summarize and encapsulate its essence. Secondly, the Deliberation objective entails encoding external graph information related to entities mentioned in the input text, utilizing a graph embedding model. Finally, the Respond objective employs a classification head model that utilizes representations from the Recap and Deliberation modules to generate the final prediction. By cascading these three models and minimizing a combined loss, we mitigate the potential for gaming the benchmark and establish a robust method for capturing the underlying semantic patterns, thus enabling accurate predictions. To evaluate the effectiveness of the RDR method, we conduct tests on multiple GLUE benchmark tasks. Our results demonstrate improved performance compared to competitive baselines, with an enhancement of up to 2\% on standard metrics. Furthermore, we analyze the observed evidence for semantic understanding exhibited by RDR models, emphasizing their ability to avoid gaming the benchmark and instead accurately capture the true underlying semantic patterns.
Authors: Shihan Dou, Enyu Zhou, Yan Liu, Songyang Gao, Jun Zhao, Wei Shen, Yuhao Zhou, Zhiheng Xi, Xiao Wang, Xiaoran Fan, Shiliang Pu, Jiang Zhu, Rui Zheng, Tao Gui, Qi Zhang, Xuanjing Huang
Abstract: Supervised fine-tuning (SFT) is a crucial step for large language models (LLMs), enabling them to align with human instructions and enhance their capabilities in downstream tasks. Increasing instruction data substantially is a direct solution to align the model with a broader range of downstream tasks or notably improve its performance on a specific task. However, we find that large-scale increases in instruction data can damage the world knowledge previously stored in LLMs. To address this challenge, we propose LoRAMoE, a novelty framework that introduces several low-rank adapters (LoRA) and integrates them by using a router network, like a plugin version of Mixture of Experts (MoE). It freezes the backbone model and forces a portion of LoRAs to focus on leveraging world knowledge to solve downstream tasks, to alleviate world knowledge-edge forgetting. Experimental results show that, as the instruction data increases, LoRAMoE can significantly improve the ability to process downstream tasks, while maintaining the world knowledge stored in the LLM.
Authors: Yuanzhi Liang, Linchao Zhu, Yi Yang
Abstract: Large Language Models (LLMs) have demonstrated their ability to replicate human behaviors across a wide range of scenarios. However, their capability in handling complex, multi-character social interactions has yet to be fully explored, primarily due to the absence of robust, quantitative evaluation methods. This gap has slowed the development of agents proficient in more nuanced interactions beyond simple exchanges, for example, small talk. To address this challenge, we introduce the Multi-Agent Interaction Evaluation Framework (AntEval), encompassing a novel interaction framework and evaluation methods. The interaction framework aims to foster an complex interaction environment that bolsters information exchange and intention expression within social interactions. Furthermore, we introduce evaluation methods, including two metrics: Information Exchanging Precision (IEP) and Interaction Expressiveness Gap (IEG), designed for the quantitative and objective assessment of agents' interaction competencies. Our findings highlight the utility of these evaluative methods and show significant potential for improving LLMs' ability to construct agents that interact in a more natural manner with human-like intricacy.
Authors: Rima Hazra, Sayan Layek, Somnath Banerjee, Soujanya Poria
Abstract: In the rapidly advancing field of artificial intelligence, the concept of Red-Teaming or Jailbreaking large language models (LLMs) has emerged as a crucial area of study. This approach is especially significant in terms of assessing and enhancing the safety and robustness of these models. This paper investigates the intricate consequences of such modifications through model editing, uncovering a complex relationship between enhancing model accuracy and preserving its ethical integrity. Our in-depth analysis reveals a striking paradox: while injecting accurate information is crucial for model reliability, it can paradoxically destabilize the model's foundational framework, resulting in unpredictable and potentially unsafe behaviors. Additionally, we propose a benchmark dataset NicheHazardQA to investigate this unsafe behavior both within the same and cross topical domain. This aspect of our research sheds light on how the edits, impact the model's safety metrics and guardrails. Our findings show that model editing serves as a cost-effective tool for topical red-teaming by methodically applying targeted edits and evaluating the resultant model behavior.
Authors: Somnath Banerjee, Amruit Sahoo, Sayan Layek, Avik Dutta, Rima Hazra, Animesh Mukherjee
Abstract: In the continuously advancing AI landscape, crafting context-rich and meaningful responses via Large Language Models (LLMs) is essential. Researchers are becoming more aware of the challenges that LLMs with fewer parameters encounter when trying to provide suitable answers to open-ended questions. To address these hurdles, the integration of cutting-edge strategies, augmentation of rich external domain knowledge to LLMs, offers significant improvements. This paper introduces a novel framework that combines graph-driven context retrieval in conjunction to knowledge graphs based enhancement, honing the proficiency of LLMs, especially in domain specific community question answering platforms like AskUbuntu, Unix, and ServerFault. We conduct experiments on various LLMs with different parameter sizes to evaluate their ability to ground knowledge and determine factual accuracy in answers to open-ended questions. Our methodology GraphContextGen consistently outperforms dominant text-based retrieval systems, demonstrating its robustness and adaptability to a larger number of use cases. This advancement highlights the importance of pairing context rich data retrieval with LLMs, offering a renewed approach to knowledge sourcing and generation in AI systems. We also show that, due to rich contextual data retrieval, the crucial entities, along with the generated answer, remain factually coherent with the gold answer.
Authors: Zhaoye Fei, Yunfan Shao, Linyang Li, Zhiyuan Zeng, Conghui He, Hang Yan, Dahua Lin, Xipeng Qiu
Abstract: Large language models have demonstrated remarkable potential in various tasks, however, there remains a significant scarcity of open-source models and data for specific domains. Previous works have primarily focused on manually specifying resources and collecting high-quality data on specific domains, which significantly consume time and effort. To address this limitation, we propose an efficient data collection method $\textit{Query of CC}$ based on large language models. This method bootstraps seed information through a large language model and retrieves related data from public corpora. It not only collects knowledge-related data for specific domains but unearths the data with potential reasoning procedures. Through the application of this method, we have curated a high-quality dataset called KNOWLEDGE PILE, encompassing four major domains, including stem and humanities sciences, among others. Experimental results demonstrate that KNOWLEDGE PILE significantly improves the performance of large language models in mathematical and knowledge-related reasoning ability tests. To facilitate academic sharing, we open-source our dataset and code, providing valuable support to the academic community.
Authors: Jung-Mei Chu, Hao-Cheng Lo, Jieh Hsiang, Chun-Chieh Cho
Abstract: In patent prosecution, timely and effective responses to Office Actions (OAs) are crucial for securing patents. However, past automation and artificial intelligence research have largely overlooked this aspect. To bridge this gap, our study introduces the Patent Office Action Response Intelligence System (PARIS) and its advanced version, the Large Language Model (LLM) Enhanced PARIS (LE-PARIS). These systems are designed to enhance the efficiency of patent attorneys in handling OA responses through collaboration with AI. The systems' key features include the construction of an OA Topics Database, development of Response Templates, and implementation of Recommender Systems and LLM-based Response Generation. To validate the effectiveness of the systems, we have employed a multi-paradigm analysis using the USPTO Office Action database and longitudinal data based on attorney interactions with our systems over six years. Through five studies, we have examined the constructiveness of OA topics (studies 1 and 2) using topic modeling and our proposed Delphi process, the efficacy of our proposed hybrid LLM-based recommender system tailored for OA responses (study 3), the quality of generated responses (study 4), and the systems' practical value in real-world scenarios through user studies (study 5). The results indicate that both PARIS and LE-PARIS significantly achieve key metrics and have a positive impact on attorney performance.
Authors: Shih-Yang Liu, Chien-Yi Wang, Hongxu Yin, Pavlo Molchanov, Yu-Chiang Frank Wang, Kwang-Ting Cheng, Min-Hung Chen
Abstract: Among the widely used parameter-efficient finetuning (PEFT) methods, LoRA and its variants have gained considerable popularity because of avoiding additional inference costs. However, there still often exists an accuracy gap between these methods and full fine-tuning (FT). In this work, we first introduce a novel weight decomposition analysis to investigate the inherent differences between FT and LoRA. Aiming to resemble the learning capacity of FT from the findings, we propose Weight-Decomposed LowRank Adaptation (DoRA). DoRA decomposes the pre-trained weight into two components, magnitude and direction, for fine-tuning, specifically employing LoRA for directional updates to efficiently minimize the number of trainable parameters. By employing DoRA, we enhance both the learning capacity and training stability of LoRA while avoiding any additional inference overhead. DoRA consistently outperforms LoRA on fine-tuning LLaMA, LLaVA, and VL-BART on various downstream tasks, such as commonsense reasoning, visual instruction tuning, and image/video-text understanding.
Authors: Yangyifei Luo, Zhuo Chen, Lingbing Guo, Qian Li, Wenxuan Zeng, Zhixin Cai, Jianxin Li
Abstract: Entity alignment (EA) aims to identify entities across different knowledge graphs that represent the same real-world objects. Recent embedding-based EA methods have achieved state-of-the-art performance in EA yet faced interpretability challenges as they purely rely on the embedding distance and neglect the logic rules behind a pair of aligned entities. In this paper, we propose the Align-Subgraph Entity Alignment (ASGEA) framework to exploit logic rules from Align-Subgraphs. ASGEA uses anchor links as bridges to construct Align-Subgraphs and spreads along the paths across KGs, which distinguishes it from the embedding-based methods. Furthermore, we design an interpretable Path-based Graph Neural Network, ASGNN, to effectively identify and integrate the logic rules across KGs. We also introduce a node-level multi-modal attention mechanism coupled with multi-modal enriched anchors to augment the Align-Subgraph. Our experimental results demonstrate the superior performance of ASGEA over the existing embedding-based methods in both EA and Multi-Modal EA (MMEA) tasks.
Authors: Hoang-Quynh Le, Duy-Cat Can, Khanh-Vinh Nguyen, Mai-Vu Tran
Abstract: This paper presents a comprehensive overview of the Comparative Opinion Mining from Vietnamese Product Reviews shared task (ComOM), held as part of the 10$^{th}$ International Workshop on Vietnamese Language and Speech Processing (VLSP 2023). The primary objective of this shared task is to advance the field of natural language processing by developing techniques that proficiently extract comparative opinions from Vietnamese product reviews. Participants are challenged to propose models that adeptly extract a comparative "quintuple" from a comparative sentence, encompassing Subject, Object, Aspect, Predicate, and Comparison Type Label. We construct a human-annotated dataset comprising $120$ documents, encompassing $7427$ non-comparative sentences and $2468$ comparisons within $1798$ sentences. Participating models undergo evaluation and ranking based on the Exact match macro-averaged quintuple F1 score.
Authors: Muhammad Maaz, Hanoona Rasheed, Abdelrahman Shaker, Salman Khan, Hisham Cholakal, Rao M. Anwer, Tim Baldwin, Michael Felsberg, Fahad S. Khan
Abstract: In pursuit of more inclusive Vision-Language Models (VLMs), this study introduces a Large Multilingual Multimodal Model called PALO. PALO offers visual reasoning capabilities in 10 major languages, including English, Chinese, Hindi, Spanish, French, Arabic, Bengali, Russian, Urdu, and Japanese, that span a total of ~5B people (65% of the world population). Our approach involves a semi-automated translation approach to adapt the multimodal instruction dataset from English to the target languages using a fine-tuned Large Language Model, thereby ensuring high linguistic fidelity while allowing scalability due to minimal manual effort. The incorporation of diverse instruction sets helps us boost overall performance across multiple languages especially those that are underrepresented like Hindi, Arabic, Bengali, and Urdu. The resulting models are trained across three scales (1.7B, 7B and 13B parameters) to show the generalization and scalability where we observe substantial improvements compared to strong baselines. We also propose the first multilingual multimodal benchmark for the forthcoming approaches to evaluate their vision-language reasoning capabilities across languages. Code: https://github.com/mbzuai-oryx/PALO.
Authors: Rodrigo Santos, Jo\~ao Silva, Lu\'is Gomes, Jo\~ao Rodrigues, Ant\'onio Branco
Abstract: To advance the neural decoding of Portuguese, in this paper we present a fully open Transformer-based, instruction-tuned decoder model that sets a new state of the art in this respect. To develop this decoder, which we named Gerv\'asio PT*, a strong LLaMA~2 7B model was used as a starting point, and its further improvement through additional training was done over language resources that include new instruction data sets of Portuguese prepared for this purpose, which are also contributed in this paper. All versions of Gerv\'asio are open source and distributed for free under an open license, including for either research or commercial usage, and can be run on consumer-grade hardware, thus seeking to contribute to the advancement of research and innovation in language technology for Portuguese.
Authors: Mihai Masala, Traian Rebedea, Horia Velicu
Abstract: In recent years,the entire field of Natural Language Processing (NLP) has enjoyed amazing novel results achieving almost human-like performance on a variety of tasks. Legal NLP domain has also been part of this process, as it has seen an impressive growth. However, general-purpose models are not readily applicable for legal domain. Due to the nature of the domain (e.g. specialized vocabulary, long documents) specific models and methods are often needed for Legal NLP. In this work we investigate both specialized and general models for predicting the final ruling of a legal case, task known as Legal Judgment Prediction (LJP). We particularly focus on methods to extend to sequence length of Transformer-based models to better understand the long documents present in legal corpora. Extensive experiments on 4 LJP datasets in Romanian, originating from 2 sources with significantly different sizes and document lengths, show that specialized models and handling long texts are critical for a good performance.
Authors: Jiantao Qiu, Haijun Lv, Zhenjiang Jin, Rui Wang, Wenchang Ning, Jia Yu, ChaoBin Zhang, Pei Chu, Yuan Qu, Jin Shi, Lindong Lu, Runyu Peng, Zhiyuan Zeng, Huanze Tang, Zhikai Lei, Jiawei Hong, Keyu Chen, Zhaoye Fei, Ruiliang Xu, Wei Li, Zhongying Tu, Hang Yan, Conghui He
Abstract: This paper presents WanJuan-CC, a safe and high-quality open-sourced English webtext dataset derived from Common Crawl data. The study addresses the challenges of constructing large-scale pre-training datasets for language models, which require vast amounts of high-quality data. A comprehensive process was designed to handle Common Crawl data, including extraction, heuristic rule filtering, fuzzy deduplication, content safety filtering, and data quality filtering. From approximately 68 billion original English documents, we obtained 2.22T Tokens of safe data and selected 1.0T Tokens of high-quality data as part of WanJuan-CC. We have open-sourced 100B Tokens from this dataset. The paper also provides statistical information related to data quality, enabling users to select appropriate data according to their needs. To evaluate the quality and utility of the dataset, we trained 1B-parameter and 3B-parameter models using WanJuan-CC and another dataset, RefinedWeb. Results show that WanJuan-CC performs better on validation datasets and downstream tasks.
Authors: Guangming Huang, Yunfei Long, Cunjin Luo, Jiaxing Shen, Xia Sun
Abstract: Pre-trained language models (PLMs) leverage chains-of-thought (CoT) to simulate human reasoning and inference processes, achieving proficient performance in multi-hop QA. However, a gap persists between PLMs' reasoning abilities and those of humans when tackling complex problems. Psychological studies suggest a vital connection between explicit information in passages and human prior knowledge during reading. Nevertheless, current research has given insufficient attention to linking input passages and PLMs' pre-training-based knowledge from the perspective of human cognition studies. In this study, we introduce a Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA. We consider the input passages as explicit knowledge, employing them to elicit implicit knowledge through unified prompt reasoning. Furthermore, our model incorporates type-specific reasoning via prompts, a form of implicit knowledge. Experimental results show that PEI performs comparably to the state-of-the-art on HotpotQA. Ablation studies confirm the efficacy of our model in bridging and integrating explicit and implicit knowledge.
Authors: Wei He, Kai Han, Yehui Tang, Chengcheng Wang, Yujie Yang, Tianyu Guo, Yunhe Wang
Abstract: Large language models (LLMs) face a daunting challenge due to the excessive computational and memory requirements of the commonly used Transformer architecture. While state space model (SSM) is a new type of foundational network architecture offering lower computational complexity, their performance has yet to fully rival that of Transformers. This paper introduces DenseSSM, a novel approach to enhance the flow of hidden information between layers in SSMs. By selectively integrating shallowlayer hidden states into deeper layers, DenseSSM retains fine-grained information crucial for the final output. Dense connections enhanced DenseSSM still maintains the training parallelizability and inference efficiency. The proposed method can be widely applicable to various SSM types like RetNet and Mamba. With similar model size, DenseSSM achieves significant improvements, exemplified by DenseRetNet outperforming the original RetNet with up to 5% accuracy improvement on public benchmarks. code is avalaible at https://github.com/WailordHe/DenseSSM
Authors: Siqi Kou, Lanxiang Hu, Zhezhi He, Zhijie Deng, Hao Zhang
Abstract: Parallel decoding methods such as Jacobi decoding show promise for more efficient LLM inference as it breaks the sequential nature of the LLM decoding process and transforms it into parallelizable computation. However, in practice, it achieves little speedup compared to traditional autoregressive (AR) decoding, primarily because Jacobi decoding seldom accurately predicts more than one token in a single fixed-point iteration step. To address this, we develop a new approach aimed at realizing fast convergence from any state to the fixed point on a Jacobi trajectory. This is accomplished by refining the target LLM to consistently predict the fixed point given any state as input. Extensive experiments demonstrate the effectiveness of our method, showing 2.4$\times$ to 3.4$\times$ improvements in generation speed while preserving generation quality across both domain-specific and open-domain benchmarks.
Authors: Sunjun Kweon, Byungjin Choi, Minkyu Kim, Rae Woong Park, Edward Choi
Abstract: We introduce KorMedMCQA, the first Korean multiple-choice question answering (MCQA) benchmark derived from Korean healthcare professional licensing examinations, covering from the year 2012 to year 2023. This dataset consists of a selection of questions from the license examinations for doctors, nurses, and pharmacists, featuring a diverse array of subjects. We conduct baseline experiments on various large language models, including proprietary/open-source, multilingual/Korean-additional pretrained, and clinical context pretrained models, highlighting the potential for further enhancements. We make our data publicly available on HuggingFace (https://huggingface.co/datasets/sean0042/KorMedMCQA) and provide a evaluation script via LM-Harness, inviting further exploration and advancement in Korean healthcare environments.
Authors: Qizhi Pei, Lijun Wu, Kaiyuan Gao, Jinhua Zhu, Yue Wang, Zun Wang, Tao Qin, Rui Yan
Abstract: The integration of biomolecular modeling with natural language (BL) has emerged as a promising interdisciplinary area at the intersection of artificial intelligence, chemistry and biology. This approach leverages the rich, multifaceted descriptions of biomolecules contained within textual data sources to enhance our fundamental understanding and enable downstream computational tasks such as biomolecule property prediction. The fusion of the nuanced narratives expressed through natural language with the structural and functional specifics of biomolecules described via various molecular modeling techniques opens new avenues for comprehensively representing and analyzing biomolecules. By incorporating the contextual language data that surrounds biomolecules into their modeling, BL aims to capture a holistic view encompassing both the symbolic qualities conveyed through language as well as quantitative structural characteristics. In this review, we provide an extensive analysis of recent advancements achieved through cross modeling of biomolecules and natural language. (1) We begin by outlining the technical representations of biomolecules employed, including sequences, 2D graphs, and 3D structures. (2) We then examine in depth the rationale and key objectives underlying effective multi-modal integration of language and molecular data sources. (3) We subsequently survey the practical applications enabled to date in this developing research area. (4) We also compile and summarize the available resources and datasets to facilitate future work. (5) Looking ahead, we identify several promising research directions worthy of further exploration and investment to continue advancing the field. The related resources and contents are updating in \url{https://github.com/QizhiPei/Awesome-Biomolecule-Language-Cross-Modeling}.
URLs: https://github.com/QizhiPei/Awesome-Biomolecule-Language-Cross-Modeling
Authors: Shiqi Chen, Miao Xiong, Junteng Liu, Zhengxuan Wu, Teng Xiao, Siyang Gao, Junxian He
Abstract: Large language models (LLMs) frequently hallucinate and produce factual errors, yet our understanding of why they make these errors remains limited. In this study, we delve into the underlying mechanisms of LLM hallucinations from the perspective of inner representations, and discover a salient pattern associated with hallucinations: correct generations tend to have sharper context activations in the hidden states of the in-context tokens, compared to the incorrect ones. Leveraging this insight, we propose an entropy-based metric to quantify the ``sharpness'' among the in-context hidden states and incorporate it into the decoding process to formulate a constrained decoding approach. Experiments on various knowledge-seeking and hallucination benchmarks demonstrate our approach's consistent effectiveness, for example, achieving up to an 8.6 point improvement on TruthfulQA. We believe this study can improve our understanding of hallucinations and serve as a practical solution for hallucination mitigation.
Authors: Nguyen Quang Duc, Le Hai Son, Nguyen Duc Nhan, Nguyen Dich Nhat Minh, Le Thanh Huong, Dinh Viet Sang
Abstract: This paper presents our contributions towards advancing the state of Vietnamese language understanding and generation through the development and dissemination of open datasets and pre-trained models for Vietnamese Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs).
Authors: Lizhou Fan, Wenyue Hua, Xiang Li, Kaijie Zhu, Mingyu Jin, Lingyao Li, Haoyang Ling, Jinkui Chi, Jindong Wang, Xin Ma, Yongfeng Zhang
Abstract: Understanding the reasoning capabilities of Multimodal Large Language Models (MLLMs) is an important area of research. In this study, we introduce a dynamic benchmark, NPHardEval4V, aimed at addressing the existing gaps in evaluating the pure reasoning abilities of MLLMs. Our benchmark aims to provide a venue to disentangle the effect of various factors such as image recognition and instruction following, from the overall performance of the models, allowing us to focus solely on evaluating their reasoning abilities. It is built by converting textual description of questions from NPHardEval to image representations. Our findings reveal significant discrepancies in reasoning abilities across different models and highlight the relatively weak performance of MLLMs compared to LLMs in terms of reasoning. We also investigate the impact of different prompting styles, including visual, text, and combined visual and text prompts, on the reasoning abilities of MLLMs, demonstrating the different impacts of multimodal inputs in model performance. Unlike traditional benchmarks, which focus primarily on static evaluations, our benchmark will be updated monthly to prevent overfitting and ensure a more authentic and fine-grained evaluation of the models. We believe that this benchmark can aid in understanding and guide the further development of reasoning abilities in MLLMs. The benchmark dataset and code are available at https://github.com/lizhouf/NPHardEval4V
Authors: Rodrigo Santos, Jo\~ao Rodrigues, Lu\'is Gomes, Jo\~ao Silva, Ant\'onio Branco, Henrique Lopes Cardoso, Tom\'as Freitas Os\'orio, Bernardo Leite
Abstract: To foster the neural encoding of Portuguese, this paper contributes foundation encoder models that represent an expansion of the still very scarce ecosystem of large language models specifically developed for this language that are fully open, in the sense that they are open source and openly distributed for free under an open license for any purpose, thus including research and commercial usages. Like most languages other than English, Portuguese is low-resourced in terms of these foundational language resources, there being the inaugural 900 million parameter Albertina and 335 million Bertimbau. Taking this couple of models as an inaugural set, we present the extension of the ecosystem of state-of-the-art open encoders for Portuguese with a larger, top performance-driven model with 1.5 billion parameters, and a smaller, efficiency-driven model with 100 million parameters. While achieving this primary goal, further results that are relevant for this ecosystem were obtained as well, namely new datasets for Portuguese based on the SuperGLUE benchmark, which we also distribute openly.
Authors: Nick Baumann, Alexander Brinkmann, Christian Bizer
Abstract: Product offers on e-commerce websites often consist of a textual product title and a textual product description. In order to provide features such as faceted product filtering or content-based product recommendation, the websites need to extract attribute-value pairs from the unstructured product descriptions. This paper explores the potential of using large language models (LLMs), such as OpenAI's GPT-3.5 and GPT-4, to extract and normalize attribute values from product titles and product descriptions. For our experiments, we introduce the WDC Product Attribute-Value Extraction (WDC PAVE) dataset. WDC PAVE consists of product offers from 87 websites that provide schema$.$org annotations. The offers belong to five different categories, each featuring a specific set of attributes. The dataset provides manually verified attribute-value pairs in two forms: (i) directly extracted values and (ii) normalized attribute values. The normalization of the attribute values requires systems to perform the following types of operations: name expansion, generalization, unit of measurement normalization, and string wrangling. Our experiments demonstrate that GPT-4 outperforms PLM-based extraction methods by 10%, achieving an F1-Score of 91%. For the extraction and normalization of product attribute values, GPT-4 achieves a similar performance to the extraction scenario, while being particularly strong at string wrangling and name expansion.
Authors: Hong Liu, Zhiyuan Li, David Hall, Percy Liang, Tengyu Ma
Abstract: Given the massive cost of language model pre-training, a non-trivial improvement of the optimization algorithm would lead to a material reduction on the time and cost of training. Adam and its variants have been state-of-the-art for years, and more sophisticated second-order (Hessian-based) optimizers often incur too much per-step overhead. In this paper, we propose Sophia, Second-order Clipped Stochastic Optimization, a simple scalable second-order optimizer that uses a light-weight estimate of the diagonal Hessian as the pre-conditioner. The update is the moving average of the gradients divided by the moving average of the estimated Hessian, followed by element-wise clipping. The clipping controls the worst-case update size and tames the negative impact of non-convexity and rapid change of Hessian along the trajectory. Sophia only estimates the diagonal Hessian every handful of iterations, which has negligible average per-step time and memory overhead. On language modeling with GPT models of sizes ranging from 125M to 1.5B, Sophia achieves a 2x speed-up compared to Adam in the number of steps, total compute, and wall-clock time, achieving the same perplexity with 50% fewer steps, less total compute, and reduced wall-clock time. Theoretically, we show that Sophia, in a much simplified setting, adapts to the heterogeneous curvatures in different parameter dimensions, and thus has a run-time bound that does not depend on the condition number of the loss.
Authors: Shengbang Tong, Erik Jones, Jacob Steinhardt
Abstract: Deployed multimodal systems can fail in ways that evaluators did not anticipate. In order to find these failures before deployment, we introduce MultiMon, a system that automatically identifies systematic failures -- generalizable, natural-language descriptions of patterns of model failures. To uncover systematic failures, MultiMon scrapes a corpus for examples of erroneous agreement: inputs that produce the same output, but should not. It then prompts a language model (e.g., GPT-4) to find systematic patterns of failure and describe them in natural language. We use MultiMon to find 14 systematic failures (e.g., "ignores quantifiers") of the CLIP text-encoder, each comprising hundreds of distinct inputs (e.g., "a shelf with a few/many books"). Because CLIP is the backbone for most state-of-the-art multimodal systems, these inputs produce failures in Midjourney 5.1, DALL-E, VideoFusion, and others. MultiMon can also steer towards failures relevant to specific use cases, such as self-driving cars. We see MultiMon as a step towards evaluation that autonomously explores the long tail of potential system failures. Code for MULTIMON is available at https://github.com/tsb0601/MultiMon.
Authors: Xuansheng Wu, Huachi Zhou, Yucheng Shi, Wenlin Yao, Xiao Huang, Ninghao Liu
Abstract: Recommendation systems help users find matched items based on their previous behaviors. Personalized recommendation becomes challenging in the absence of historical user-item interactions, a practical problem for startups known as the system cold-start recommendation. While existing research addresses cold-start issues for either users or items, we still lack solutions for system cold-start scenarios. To tackle the problem, we propose PromptRec, a simple but effective approach based on in-context learning of language models, where we transform the recommendation task into the sentiment analysis task on natural language containing user and item profiles. However, this naive approach heavily relies on the strong in-context learning ability emerged from large language models, which could suffer from significant latency for online recommendations. To solve the challenge, we propose to enhance small language models for recommender systems with a data-centric pipeline, which consists of: (1) constructing a refined corpus for model pre-training; (2) constructing a decomposed prompt template via prompt pre-training. They correspond to the development of training data and inference data, respectively. The pipeline is supported by a theoretical framework that formalizes the connection between in-context recommendation and language modeling. To evaluate our approach, we introduce a cold-start recommendation benchmark, and the results demonstrate that the enhanced small language models can achieve comparable cold-start recommendation performance to that of large models with only $17\%$ of the inference time. To the best of our knowledge, this is the first study to tackle the system cold-start recommendation problem. We believe our findings will provide valuable insights for future works. The benchmark and implementations are available at https://github.com/JacksonWuxs/PromptRec.
Authors: Shuijing Liu, Aamir Hasan, Kaiwen Hong, Runxuan Wang, Peixin Chang, Zachary Mizrachi, Justin Lin, D. Livingston McPherson, Wendy A. Rogers, Katherine Driggs-Campbell
Abstract: Persons with visual impairments (PwVI) have difficulties understanding and navigating spaces around them. Current wayfinding technologies either focus solely on navigation or provide limited communication about the environment. Motivated by recent advances in visual-language grounding and semantic navigation, we propose DRAGON, a guiding robot powered by a dialogue system and the ability to associate the environment with natural language. By understanding the commands from the user, DRAGON is able to guide the user to the desired landmarks on the map, describe the environment, and answer questions from visual observations. Through effective utilization of dialogue, the robot can ground the user's free-form descriptions to landmarks in the environment, and give the user semantic information through spoken language. We conduct a user study with blindfolded participants in an everyday indoor environment. Our results demonstrate that DRAGON is able to communicate with the user smoothly, provide a good guiding experience, and connect users with their surrounding environment in an intuitive manner. Videos and code are available at https://sites.google.com/view/dragon-wayfinding/home.
Authors: Guilherme F. C. F. Almeida, Jos\'e Luiz Nunes, Neele Engelmann, Alex Wiegmann, Marcelo de Ara\'ujo
Abstract: Large language models (LLMs) exhibit expert-level performance in tasks across a wide range of different domains. Ethical issues raised by LLMs and the need to align future versions makes it important to know how state of the art models reason about moral and legal issues. In this paper, we employ the methods of experimental psychology to probe into this question. We replicate eight studies from the experimental literature with instances of Google's Gemini Pro, Anthropic's Claude 2.1, OpenAI's GPT-4, and Meta's Llama 2 Chat 70b. We find that alignment with human responses shifts from one experiment to another, and that models differ amongst themselves as to their overall alignment, with GPT-4 taking a clear lead over all other models we tested. Nonetheless, even when LLM-generated responses are highly correlated to human responses, there are still systematic differences, with a tendency for models to exaggerate effects that are present among humans, in part by reducing variance. This recommends caution with regards to proposals of replacing human participants with current state-of-the-art LLMs in psychological research and highlights the need for further research about the distinctive aspects of machine psychology.
Authors: Zhaowei Li, Qi Xu, Dong Zhang, Hang Song, Yiqing Cai, Qi Qi, Ran Zhou, Junting Pan, Zefeng Li, Van Tu Vu, Zhida Huang, Tao Wang
Abstract: Multi-modal large language models have demonstrated impressive performance across various tasks in different modalities. However, existing multi-modal models primarily emphasize capturing global information within each modality while neglecting the importance of perceiving local information across modalities. Consequently, these models lack the ability to effectively understand the fine-grained details of input data, limiting their performance in tasks that require a more nuanced understanding. To address this limitation, there is a compelling need to develop models that enable fine-grained understanding across multiple modalities, thereby enhancing their applicability to a wide range of tasks. In this paper, we propose GroundingGPT, a language enhanced multi-modal grounding model. Beyond capturing global information like other multi-modal models, our proposed model excels at tasks demanding a detailed understanding of local information within the input. It demonstrates precise identification and localization of specific regions in images or moments in videos. To achieve this objective, we design a diversified dataset construction pipeline, resulting in a multi-modal, multi-granularity dataset for model training. The code, dataset, and demo of our model can be found at https: //github.com/lzw-lzw/GroundingGPT.
Authors: Jia Li, Ge Li, Yunfei Zhao, Yongmin Li, Zhi Jin, Hao Zhu, Huanyu Liu, Kaibo Liu, Lecheng Wang, Zheng Fang, Lanshen Wang, Jiazheng Ding, Xuanming Zhang, Yihong Dong, Yuqi Zhu, Bin Gu, Mengfei Yang
Abstract: How to evaluate Large Language Models (LLMs) in code generation is an open question. Many benchmarks have been proposed but are inconsistent with practical software projects, e.g., unreal program distributions, insufficient dependencies, and small-scale project contexts. Thus, the capabilities of LLMs in practical projects are still unclear. In this paper, we propose a new benchmark named DevEval, aligned with Developers' experiences in practical projects. DevEval is collected through a rigorous pipeline, containing 2,690 samples from 119 practical projects and covering 10 domains. Compared to previous benchmarks, DevEval aligns to practical projects in multiple dimensions, e.g., real program distributions, sufficient dependencies, and enough-scale project contexts. We assess five popular LLMs on DevEval (e.g., gpt-4, gpt-3.5-turbo, CodeLLaMa, and StarCoder) and reveal their actual abilities in code generation. For instance, the highest Pass@1 of gpt-3.5-turbo only is 42 in our experiments. We also discuss the challenges and future directions of code generation in practical projects. We open-source DevEval and hope it can facilitate the development of code generation in practical projects.
Authors: Naoyuki Kanda, Xiaofei Wang, Sefik Emre Eskimez, Manthan Thakker, Hemin Yang, Zirun Zhu, Min Tang, Canrun Li, Chung-Hsien Tsai, Zhen Xiao, Yufei Xia, Jinzhu Li, Yanqing Liu, Sheng Zhao, Michael Zeng
Abstract: Laughter is one of the most expressive and natural aspects of human speech, conveying emotions, social cues, and humor. However, most text-to-speech (TTS) systems lack the ability to produce realistic and appropriate laughter sounds, limiting their applications and user experience. While there have been prior works to generate natural laughter, they fell short in terms of controlling the timing and variety of the laughter to be generated. In this work, we propose ELaTE, a zero-shot TTS that can generate natural laughing speech of any speaker based on a short audio prompt with precise control of laughter timing and expression. Specifically, ELaTE works on the audio prompt to mimic the voice characteristic, the text prompt to indicate the contents of the generated speech, and the input to control the laughter expression, which can be either the start and end times of laughter, or the additional audio prompt that contains laughter to be mimicked. We develop our model based on the foundation of conditional flow-matching-based zero-shot TTS, and fine-tune it with frame-level representation from a laughter detector as additional conditioning. With a simple scheme to mix small-scale laughter-conditioned data with large-scale pre-training data, we demonstrate that a pre-trained zero-shot TTS model can be readily fine-tuned to generate natural laughter with precise controllability, without losing any quality of the pre-trained zero-shot TTS model. Through objective and subjective evaluations, we show that ELaTE can generate laughing speech with significantly higher quality and controllability compared to conventional models. See https://aka.ms/elate/ for demo samples.
URLs: https://aka.ms/elate/
Authors: Mingtian Zhang, Shawn Lan, Peter Hayes, David Barber
Abstract: Retrieval Augmented Generation (RAG) has emerged as an effective solution for mitigating hallucinations in Large Language Models (LLMs). The retrieval stage in RAG typically involves a pre-trained embedding model, which converts queries and passages into vectors to capture their semantics. However, a standard pre-trained embedding model may exhibit sub-optimal performance when applied to specific domain knowledge, necessitating fine-tuning. This paper addresses scenarios where the embeddings are only available from a black-box model. We introduce Model augmented fine-tuning (Mafin) -- a novel approach for fine-tuning a black-box embedding model by augmenting it with a trainable embedding model. Our results demonstrate that Mafin significantly enhances the performance of the black-box embeddings by only requiring the training of a small augmented model. We validate the effectiveness of our method on both labeled and unlabeled datasets, illustrating its broad applicability and efficiency.
Authors: Jizhi Zhang, Keqin Bao, Wenjie Wang, Yang Zhang, Wentao Shi, Wanhong Xu, Fuli Feng, Tat-Seng Chua
Abstract: The new kind of Agent-oriented information system, exemplified by GPTs, urges us to inspect the information system infrastructure to support Agent-level information processing and to adapt to the characteristics of Large Language Model (LLM)-based Agents, such as interactivity. In this work, we envisage the prospect of the recommender system on LLM-based Agent platforms and introduce a novel recommendation paradigm called Rec4Agentverse, comprised of Agent Items and Agent Recommender. Rec4Agentverse emphasizes the collaboration between Agent Items and Agent Recommender, thereby promoting personalized information services and enhancing the exchange of information beyond the traditional user-recommender feedback loop. Additionally, we prospect the evolution of Rec4Agentverse and conceptualize it into three stages based on the enhancement of the interaction and information exchange among Agent Items, Agent Recommender, and the user. A preliminary study involving several cases of Rec4Agentverse validates its significant potential for application. Lastly, we discuss potential issues and promising directions for future research.
Authors: Xiaomeng Hu, Pin-Yu Chen, Tsung-Yi Ho
Abstract: Large Language Models (LLMs) are becoming a prominent generative AI tool, where the user enters a query and the LLM generates an answer. To reduce harm and misuse, efforts have been made to align these LLMs to human values using advanced training techniques such as Reinforcement Learning from Human Feedback (RLHF). However, recent studies have highlighted the vulnerability of LLMs to adversarial jailbreak attempts aiming at subverting the embedded safety guardrails. To address this challenge, this paper defines and investigates the Refusal Loss of LLMs and then proposes a method called Gradient Cuff to detect jailbreak attempts. Gradient Cuff exploits the unique properties observed in the refusal loss landscape, including functional values and its smoothness, to design an effective two-step detection strategy. Experimental results on two aligned LLMs (LLaMA-2-7B-Chat and Vicuna-7B-V1.5) and six types of jailbreak attacks (GCG, AutoDAN, PAIR, TAP, Base64, and LRL) show that Gradient Cuff can significantly improve the LLM's rejection capability for malicious jailbreak queries, while maintaining the model's performance for benign user queries by adjusting the detection threshold.
Authors: Zeyuan Ma, Hongshu Guo, Jiacheng Chen, Guojun Peng, Zhiguang Cao, Yining Ma, Yue-Jiao Gong
Abstract: Recent research explores optimization using large language models (LLMs) by either iteratively seeking next-step solutions from LLMs or directly prompting LLMs for an optimizer. However, these approaches exhibit inherent limitations, including low operational efficiency, high sensitivity to prompt design, and a lack of domain-specific knowledge. We introduce LLaMoCo, the first instruction-tuning framework designed to adapt LLMs for solving optimization problems in a code-to-code manner. Specifically, we establish a comprehensive instruction set containing well-described problem prompts and effective optimization codes. We then develop a novel two-phase learning strategy that incorporates a contrastive learning-based warm-up procedure before the instruction-tuning phase to enhance the convergence behavior during model fine-tuning. The experiment results demonstrate that a CodeGen (350M) model fine-tuned by our LLaMoCo achieves superior optimization performance compared to GPT-4 Turbo and the other competitors across both synthetic and realistic problem sets. The fine-tuned model and the usage instructions are available at https://anonymous.4open.science/r/LLaMoCo-722A.