Title
stringlengths 16
196
| Authors
stringlengths 6
6.27k
| Abstract
stringlengths 242
1.92k
| entry_id
stringlengths 33
33
| Date
unknown | Categories
stringclasses 597
values | year
int32 2.02k
2.02k
|
|---|---|---|---|---|---|---|
DeepSpeed-VisualChat: Multi-Round Multi-Image Interleave Chat via Multi-Modal Causal Attention | Zhewei Yao, Xiaoxia Wu, Conglong Li, Minjia Zhang, Heyang Qin, Olatunji Ruwase, Ammar Ahmad Awan, Samyam Rajbhandari, Yuxiong He | Most of the existing multi-modal models, hindered by their incapacity to adeptly manage interleaved image-and-text inputs in multi-image, multi-round dialogues, face substantial constraints in resource allocation for training and data accessibility, impacting their adaptability and scalability across varied interaction realms. To address this, we present the DeepSpeed-VisualChat framework, designed to optimize Large Language Models (LLMs) by incorporating multi-modal capabilities, with a focus on enhancing the proficiency of Large Vision and Language Models in handling interleaved inputs. Our framework is notable for (1) its open-source support for multi-round and multi-image dialogues, (2) introducing an innovative multi-modal causal attention mechanism, and (3) utilizing data blending techniques on existing datasets to assure seamless interactions in multi-round, multi-image conversations. Compared to existing frameworks, DeepSpeed-VisualChat shows superior scalability up to 70B parameter language model size, representing a significant advancement in multi-modal language models and setting a solid foundation for future explorations. | http://arxiv.org/abs/2309.14327v3 | "2023-09-25T17:53:29Z" | cs.CV, cs.CL | 2,023 |
Physics of Language Models: Part 3.1, Knowledge Storage and Extraction | Zeyuan Allen-Zhu, Yuanzhi Li | Large language models (LLMs) can store a vast amount of world knowledge, often extractable via question-answering (e.g., "What is Abraham Lincoln's birthday?"). However, do they answer such questions based on exposure to similar questions during training (i.e., cheating), or by genuinely learning to extract knowledge from sources like Wikipedia? In this paper, we investigate this issue using a controlled biography dataset. We find a strong correlation between the model's ability to extract knowledge and various diversity measures of the training data. $\textbf{Essentially}$, for knowledge to be reliably extracted, it must be sufficiently augmented (e.g., through paraphrasing, sentence shuffling) $\textit{during pretraining}$. Without such augmentation, knowledge may be memorized but not extractable, leading to 0% accuracy, regardless of subsequent instruction fine-tuning. To understand why this occurs, we employ (nearly) linear probing to demonstrate a strong connection between the observed correlation and how the model internally encodes knowledge -- whether it is linearly encoded in the hidden embeddings of entity names or distributed across other token embeddings in the training text. This paper provides $\textbf{several key recommendations for LLM pretraining in the industry}$: (1) rewrite the pretraining data -- using small, auxiliary models -- to provide knowledge augmentation, and (2) incorporate more instruction-finetuning data into the pretraining stage before it becomes too late. | http://arxiv.org/abs/2309.14316v2 | "2023-09-25T17:37:20Z" | cs.CL, cs.AI, cs.LG | 2,023 |
Connecting Speech Encoder and Large Language Model for ASR | Wenyi Yu, Changli Tang, Guangzhi Sun, Xianzhao Chen, Tian Tan, Wei Li, Lu Lu, Zejun Ma, Chao Zhang | The impressive capability and versatility of large language models (LLMs) have aroused increasing attention in automatic speech recognition (ASR), with several pioneering studies attempting to build integrated ASR models by connecting a speech encoder with an LLM. This paper presents a comparative study of three commonly used structures as connectors, including fully connected layers, multi-head cross-attention, and Q-Former. Speech encoders from the Whisper model series as well as LLMs from the Vicuna model series with different model sizes were studied. Experiments were performed on the commonly used LibriSpeech, Common Voice, and GigaSpeech datasets, where the LLMs with Q-Formers demonstrated consistent and considerable word error rate (WER) reductions over LLMs with other connector structures. Q-Former-based LLMs can generalise well to out-of-domain datasets, where 12% relative WER reductions over the Whisper baseline ASR model were achieved on the Eval2000 test set without using any in-domain training data from Switchboard. Moreover, a novel segment-level Q-Former is proposed to enable LLMs to recognise speech segments with a duration exceeding the limitation of the encoders, which results in 17% relative WER reductions over other connector structures on 90-second-long speech data. | http://arxiv.org/abs/2309.13963v2 | "2023-09-25T08:57:07Z" | eess.AS, cs.CL, cs.SD | 2,023 |
Analyzing the Efficacy of an LLM-Only Approach for Image-based Document Question Answering | Nidhi Hegde, Sujoy Paul, Gagan Madan, Gaurav Aggarwal | Recent document question answering models consist of two key components: the vision encoder, which captures layout and visual elements in images, and a Large Language Model (LLM) that helps contextualize questions to the image and supplements them with external world knowledge to generate accurate answers. However, the relative contributions of the vision encoder and the language model in these tasks remain unclear. This is especially interesting given the effectiveness of instruction-tuned LLMs, which exhibit remarkable adaptability to new tasks. To this end, we explore the following aspects in this work: (1) The efficacy of an LLM-only approach on document question answering tasks (2) strategies for serializing textual information within document images and feeding it directly to an instruction-tuned LLM, thus bypassing the need for an explicit vision encoder (3) thorough quantitative analysis on the feasibility of such an approach. Our comprehensive analysis encompasses six diverse benchmark datasets, utilizing LLMs of varying scales. Our findings reveal that a strategy exclusively reliant on the LLM yields results that are on par with or closely approach state-of-the-art performance across a range of datasets. We posit that this evaluation framework will serve as a guiding resource for selecting appropriate datasets for future research endeavors that emphasize the fundamental importance of layout and image content information. | http://arxiv.org/abs/2309.14389v1 | "2023-09-25T07:01:16Z" | cs.CV, cs.AI | 2,023 |
AutoPrep: An Automatic Preprocessing Framework for In-the-Wild Speech Data | Jianwei Yu, Hangting Chen, Yanyao Bian, Xiang Li, Yi Luo, Jinchuan Tian, Mengyang Liu, Jiayi Jiang, Shuai Wang | Recently, the utilization of extensive open-sourced text data has significantly advanced the performance of text-based large language models (LLMs). However, the use of in-the-wild large-scale speech data in the speech technology community remains constrained. One reason for this limitation is that a considerable amount of the publicly available speech data is compromised by background noise, speech overlapping, lack of speech segmentation information, missing speaker labels, and incomplete transcriptions, which can largely hinder their usefulness. On the other hand, human annotation of speech data is both time-consuming and costly. To address this issue, we introduce an automatic in-the-wild speech data preprocessing framework (AutoPrep) in this paper, which is designed to enhance speech quality, generate speaker labels, and produce transcriptions automatically. The proposed AutoPrep framework comprises six components: speech enhancement, speech segmentation, speaker clustering, target speech extraction, quality filtering and automatic speech recognition. Experiments conducted on the open-sourced WenetSpeech and our self-collected AutoPrepWild corpora demonstrate that the proposed AutoPrep framework can generate preprocessed data with similar DNSMOS and PDNSMOS scores compared to several open-sourced TTS datasets. The corresponding TTS system can achieve up to 0.68 in-domain speaker similarity. | http://arxiv.org/abs/2309.13905v1 | "2023-09-25T07:01:10Z" | eess.AS, cs.SD | 2,023 |
Can LLM-Generated Misinformation Be Detected? | Canyu Chen, Kai Shu | The advent of Large Language Models (LLMs) has made a transformative impact. However, the potential that LLMs such as ChatGPT can be exploited to generate misinformation has posed a serious concern to online safety and public trust. A fundamental research question is: will LLM-generated misinformation cause more harm than human-written misinformation? We propose to tackle this question from the perspective of detection difficulty. We first build a taxonomy of LLM-generated misinformation. Then we categorize and validate the potential real-world methods for generating misinformation with LLMs. Then, through extensive empirical investigation, we discover that LLM-generated misinformation can be harder to detect for humans and detectors compared to human-written misinformation with the same semantics, which suggests it can have more deceptive styles and potentially cause more harm. We also discuss the implications of our discovery on combating misinformation in the age of LLMs and the countermeasures. | http://arxiv.org/abs/2309.13788v5 | "2023-09-25T00:45:07Z" | cs.CL, cs.AI, cs.CR, cs.HC, cs.LG | 2,023 |
MentaLLaMA: Interpretable Mental Health Analysis on Social Media with Large Language Models | Kailai Yang, Tianlin Zhang, Ziyan Kuang, Qianqian Xie, Jimin Huang, Sophia Ananiadou | With the development of web technology, social media texts are becoming a rich source for automatic mental health analysis. As traditional discriminative methods bear the problem of low interpretability, the recent large language models have been explored for interpretable mental health analysis on social media, which aims to provide detailed explanations along with predictions. The results show that ChatGPT can generate approaching-human explanations for its correct classifications. However, LLMs still achieve unsatisfactory classification performance in a zero-shot/few-shot manner. Domain-specific finetuning is an effective solution, but faces 2 challenges: 1) lack of high-quality training data. 2) no open-source LLMs for interpretable mental health analysis were released to lower the finetuning cost. To alleviate these problems, we build the first multi-task and multi-source interpretable mental health instruction (IMHI) dataset on social media, with 105K data samples. The raw social media data are collected from 10 existing sources covering 8 mental health analysis tasks. We use expert-written few-shot prompts and collected labels to prompt ChatGPT and obtain explanations from its responses. To ensure the reliability of the explanations, we perform strict automatic and human evaluations on the correctness, consistency, and quality of generated data. Based on the IMHI dataset and LLaMA2 foundation models, we train MentalLLaMA, the first open-source LLM series for interpretable mental health analysis with instruction-following capability. We also evaluate the performance of MentalLLaMA on the IMHI evaluation benchmark with 10 test sets, where their correctness for making predictions and the quality of explanations are examined. The results show that MentalLLaMA approaches state-of-the-art discriminative methods in correctness and generates high-quality explanations. | http://arxiv.org/abs/2309.13567v3 | "2023-09-24T06:46:08Z" | cs.CL | 2,023 |
Enhancing Student Performance Prediction on Learnersourced Questions with SGNN-LLM Synergy | Lin Ni, Sijie Wang, Zeyu Zhang, Xiaoxuan Li, Xianda Zheng, Paul Denny, Jiamou Liu | Learnersourcing offers great potential for scalable education through student content creation. However, predicting student performance on learnersourced questions, which is essential for personalizing the learning experience, is challenging due to the inherent noise in student-generated data. Moreover, while conventional graph-based methods can capture the complex network of student and question interactions, they often fall short under cold start conditions where limited student engagement with questions yields sparse data. To address both challenges, we introduce an innovative strategy that synergizes the potential of integrating Signed Graph Neural Networks (SGNNs) and Large Language Model (LLM) embeddings. Our methodology employs a signed bipartite graph to comprehensively model student answers, complemented by a contrastive learning framework that enhances noise resilience. Furthermore, LLM's contribution lies in generating foundational question embeddings, proving especially advantageous in addressing cold start scenarios characterized by limited graph data. Validation across five real-world datasets sourced from the PeerWise platform underscores our approach's effectiveness. Our method outperforms baselines, showcasing enhanced predictive accuracy and robustness. | http://arxiv.org/abs/2309.13500v3 | "2023-09-23T23:37:55Z" | cs.LG, cs.AI, 97P80 | 2,023 |
Resolving References in Visually-Grounded Dialogue via Text Generation | Bram Willemsen, Livia Qian, Gabriel Skantze | Vision-language models (VLMs) have shown to be effective at image retrieval based on simple text queries, but text-image retrieval based on conversational input remains a challenge. Consequently, if we want to use VLMs for reference resolution in visually-grounded dialogue, the discourse processing capabilities of these models need to be augmented. To address this issue, we propose fine-tuning a causal large language model (LLM) to generate definite descriptions that summarize coreferential information found in the linguistic context of references. We then use a pretrained VLM to identify referents based on the generated descriptions, zero-shot. We evaluate our approach on a manually annotated dataset of visually-grounded dialogues and achieve results that, on average, exceed the performance of the baselines we compare against. Furthermore, we find that using referent descriptions based on larger context windows has the potential to yield higher returns. | http://arxiv.org/abs/2309.13430v1 | "2023-09-23T17:07:54Z" | cs.CL, cs.AI, cs.CV | 2,023 |
BAMBOO: A Comprehensive Benchmark for Evaluating Long Text Modeling Capacities of Large Language Models | Zican Dong, Tianyi Tang, Junyi Li, Wayne Xin Zhao, Ji-Rong Wen | Large language models (LLMs) have achieved dramatic proficiency over NLP tasks with normal length. Recently, multiple studies have committed to extending the context length and enhancing the long text modeling capabilities of LLMs. To comprehensively evaluate the long context ability of LLMs, we propose BAMBOO, a multi-task long context benchmark. BAMBOO has been designed with four principles: comprehensive capacity evaluation, avoidance of data contamination, accurate automatic evaluation, and different length levels. It consists of 10 datasets from 5 different long text understanding tasks, i.e. question answering, hallucination detection, text sorting, language modeling, and code completion, to cover core capacities and various domains of LLMs. We conduct experiments with five long context models on BAMBOO and further discuss four key research questions of long text. We also qualitatively analyze current long context models and point out future directions for enhancing long text modeling capacities. We release our data, prompts, and code at https://github.com/RUCAIBox/BAMBOO. | http://arxiv.org/abs/2309.13345v3 | "2023-09-23T11:36:15Z" | cs.CL | 2,023 |
Towards LLM-guided Causal Explainability for Black-box Text Classifiers | Amrita Bhattacharjee, Raha Moraffah, Joshua Garland, Huan Liu | With the advent of larger and more complex deep learning models, such as in Natural Language Processing (NLP), model qualities like explainability and interpretability, albeit highly desirable, are becoming harder challenges to tackle and solve. For example, state-of-the-art models in text classification are black-box by design. Although standard explanation methods provide some degree of explainability, these are mostly correlation-based methods and do not provide much insight into the model. The alternative of causal explainability is more desirable to achieve but extremely challenging in NLP due to a variety of reasons. Inspired by recent endeavors to utilize Large Language Models (LLMs) as experts, in this work, we aim to leverage the instruction-following and textual understanding capabilities of recent state-of-the-art LLMs to facilitate causal explainability via counterfactual explanation generation for black-box text classifiers. To do this, we propose a three-step pipeline via which, we use an off-the-shelf LLM to: (1) identify the latent or unobserved features in the input text, (2) identify the input features associated with the latent features, and finally (3) use the identified input features to generate a counterfactual explanation. We experiment with our pipeline on multiple NLP text classification datasets, with several recent LLMs, and present interesting and promising findings. | http://arxiv.org/abs/2309.13340v2 | "2023-09-23T11:22:28Z" | cs.CL, cs.AI, cs.LG | 2,023 |
Calibrating LLM-Based Evaluator | Yuxuan Liu, Tianchi Yang, Shaohan Huang, Zihan Zhang, Haizhen Huang, Furu Wei, Weiwei Deng, Feng Sun, Qi Zhang | Recent advancements in large language models (LLMs) on language modeling and emergent capabilities make them a promising reference-free evaluator of natural language generation quality, and a competent alternative to human evaluation. However, hindered by the closed-source or high computational demand to host and tune, there is a lack of practice to further calibrate an off-the-shelf LLM-based evaluator towards better human alignment. In this work, we propose AutoCalibrate, a multi-stage, gradient-free approach to automatically calibrate and align an LLM-based evaluator toward human preference. Instead of explicitly modeling human preferences, we first implicitly encompass them within a set of human labels. Then, an initial set of scoring criteria is drafted by the language model itself, leveraging in-context learning on different few-shot examples. To further calibrate this set of criteria, we select the best performers and re-draft them with self-refinement. Our experiments on multiple text quality evaluation datasets illustrate a significant improvement in correlation with expert evaluation through calibration. Our comprehensive qualitative analysis conveys insightful intuitions and observations on the essence of effective scoring criteria. | http://arxiv.org/abs/2309.13308v1 | "2023-09-23T08:46:11Z" | cs.CL | 2,023 |
User Simulation with Large Language Models for Evaluating Task-Oriented Dialogue | Sam Davidson, Salvatore Romeo, Raphael Shu, James Gung, Arshit Gupta, Saab Mansour, Yi Zhang | One of the major impediments to the development of new task-oriented dialogue (TOD) systems is the need for human evaluation at multiple stages and iterations of the development process. In an effort to move toward automated evaluation of TOD, we propose a novel user simulator built using recently developed large pretrained language models (LLMs). In order to increase the linguistic diversity of our system relative to the related previous work, we do not fine-tune the LLMs used by our system on existing TOD datasets; rather we use in-context learning to prompt the LLMs to generate robust and linguistically diverse output with the goal of simulating the behavior of human interlocutors. Unlike previous work, which sought to maximize goal success rate (GSR) as the primary metric of simulator performance, our goal is a system which achieves a GSR similar to that observed in human interactions with TOD systems. Using this approach, our current simulator is effectively able to interact with several TOD systems, especially on single-intent conversational goals, while generating lexically and syntactically diverse output relative to previous simulators that rely upon fine-tuned models. Finally, we collect a Human2Bot dataset of humans interacting with the same TOD systems with which we experimented in order to better quantify these achievements. | http://arxiv.org/abs/2309.13233v1 | "2023-09-23T02:04:57Z" | cs.CL | 2,023 |
Investigating Large Language Models and Control Mechanisms to Improve Text Readability of Biomedical Abstracts | Zihao Li, Samuel Belkadi, Nicolo Micheletti, Lifeng Han, Matthew Shardlow, Goran Nenadic | Biomedical literature often uses complex language and inaccessible professional terminologies. That is why simplification plays an important role in improving public health literacy. Applying Natural Language Processing (NLP) models to automate such tasks allows for quick and direct accessibility for lay readers. In this work, we investigate the ability of state-of-the-art large language models (LLMs) on the task of biomedical abstract simplification, using the publicly available dataset for plain language adaptation of biomedical abstracts (\textbf{PLABA}). The methods applied include domain fine-tuning and prompt-based learning (PBL) on: 1) Encoder-decoder models (T5, SciFive, and BART), 2) Decoder-only GPT models (GPT-3.5 and GPT-4) from OpenAI and BioGPT, and 3) Control-token mechanisms on BART-based models. We used a range of automatic evaluation metrics, including BLEU, ROUGE, SARI, and BERTscore, and also conducted human evaluations. BART-Large with Control Token (BART-L-w-CT) mechanisms reported the highest SARI score of 46.54 and T5-base reported the highest BERTscore 72.62. In human evaluation, BART-L-w-CTs achieved a better simplicity score over T5-Base (2.9 vs. 2.2), while T5-Base achieved a better meaning preservation score over BART-L-w-CTs (3.1 vs. 2.6). We also categorised the system outputs with examples, hoping this will shed some light for future research on this task. Our code, fine-tuned models, and data splits are available at \url{https://github.com/HECTA-UoM/PLABA-MU} \begin{IEEEkeywords} Large Language Models, Text Simplification, Biomedical NLP, Control Mechanisms, Health Informatics \end{IEEEkeywords} | http://arxiv.org/abs/2309.13202v2 | "2023-09-22T22:47:32Z" | cs.CL, cs.AI | 2,023 |
Towards Green AI in Fine-tuning Large Language Models via Adaptive Backpropagation | Kai Huang, Hanyun Yin, Heng Huang, Wei Gao | Fine-tuning is the most effective way of adapting pre-trained large language models (LLMs) to downstream applications. With the fast growth of LLM-enabled AI applications and democratization of open-souced LLMs, fine-tuning has become possible for non-expert individuals, but intensively performed LLM fine-tuning worldwide could result in significantly high energy consumption and carbon footprint, which may bring large environmental impact. Mitigating such environmental impact towards Green AI directly correlates to reducing the FLOPs of fine-tuning, but existing techniques on efficient LLM fine-tuning can only achieve limited reduction of such FLOPs, due to their ignorance of the backpropagation cost in fine-tuning. To address this limitation, in this paper we present GreenTrainer, a new LLM fine-tuning technique that adaptively evaluates different tensors' backpropagation costs and contributions to the fine-tuned model accuracy, to minimize the fine-tuning cost by selecting the most appropriate set of tensors in training. Such selection in GreenTrainer is made based on a given objective of FLOPs reduction, which can flexibly adapt to the carbon footprint in energy supply and the need in Green AI. Experiment results over multiple open-sourced LLM models and abstractive summarization datasets show that, compared to fine-tuning the whole LLM model, GreenTrainer can save up to 64% FLOPs in fine-tuning without any noticeable model accuracy loss. Compared to the existing fine-tuning techniques such as LoRa, GreenTrainer can achieve up to 4% improvement on model accuracy with on-par FLOPs reduction. | http://arxiv.org/abs/2309.13192v2 | "2023-09-22T21:55:18Z" | cs.LG, cs.AI | 2,023 |
Effective Distillation of Table-based Reasoning Ability from LLMs | Bohao Yang, Chen Tang, Kun Zhao, Chenghao Xiao, Chenghua Lin | Large Language Models (LLMs) have demonstrated remarkable performance across a wide range of natural language processing tasks. However, their enormous parameter size and extremely high requirements for compute power pose challenges for their practical deployment. Recent research has revealed that specific capabilities of LLMs, such as numerical reasoning, can be transferred to smaller models through distillation. Some studies explore the potential of leveraging LLMs to perform table-based reasoning. However, there has been no prior work focusing on table reasoning skills in smaller models specifically tailored for scientific table-to-text generation tasks. In this paper, we propose a novel table-based reasoning distillation approach, with the aim of distilling LLMs into tailored smaller models. Our experimental results have shown that a 220 million parameter model (Flan-T5-base) fine-tuned using distilled data, not only achieves a significant improvement compared to traditionally fine-tuned baselines, but also surpasses specific LLMs on a scientific table-to-text generation dataset. Our code is available at https://github.com/Bernard-Yang/DistillTableCoT. | http://arxiv.org/abs/2309.13182v2 | "2023-09-22T21:15:28Z" | cs.CL | 2,023 |
Contextual Emotion Estimation from Image Captions | Vera Yang, Archita Srivastava, Yasaman Etesam, Chuxuan Zhang, Angelica Lim | Emotion estimation in images is a challenging task, typically using computer vision methods to directly estimate people's emotions using face, body pose and contextual cues. In this paper, we explore whether Large Language Models (LLMs) can support the contextual emotion estimation task, by first captioning images, then using an LLM for inference. First, we must understand: how well do LLMs perceive human emotions? And which parts of the information enable them to determine emotions? One initial challenge is to construct a caption that describes a person within a scene with information relevant for emotion perception. Towards this goal, we propose a set of natural language descriptors for faces, bodies, interactions, and environments. We use them to manually generate captions and emotion annotations for a subset of 331 images from the EMOTIC dataset. These captions offer an interpretable representation for emotion estimation, towards understanding how elements of a scene affect emotion perception in LLMs and beyond. Secondly, we test the capability of a large language model to infer an emotion from the resulting image captions. We find that GPT-3.5, specifically the text-davinci-003 model, provides surprisingly reasonable emotion predictions consistent with human annotations, but accuracy can depend on the emotion concept. Overall, the results suggest promise in the image captioning and LLM approach. | http://arxiv.org/abs/2309.13136v1 | "2023-09-22T18:44:34Z" | cs.CV, cs.AI | 2,023 |
ReConcile: Round-Table Conference Improves Reasoning via Consensus among Diverse LLMs | Justin Chih-Yao Chen, Swarnadeep Saha, Mohit Bansal | Large Language Models (LLMs) still struggle with natural language reasoning tasks. Motivated by the society of minds (Minsky, 1988), we propose ReConcile, a multi-model multiagent framework designed as a round table conference among diverse LLM agents. ReConcile enhances collaborative reasoning between LLM agents via multiple rounds of discussion, learning to convince other agents to improve their answers, and employing a confidence-weighted voting mechanism that leads to a better consensus. In each round, ReConcile initiates discussion between agents via a 'discussion prompt' that consists of (a) grouped answers and explanations generated by each agent in the previous round, (b) their confidence scores, and (c) demonstrations of answer-rectifying human explanations, used for convincing other agents. Experiments on seven benchmarks demonstrate that ReConcile significantly improves LLMs' reasoning -- both individually and as a team -- surpassing prior single-agent and multi-agent baselines by up to 11.4% and even outperforming GPT-4 on three datasets. ReConcile also flexibly incorporates different combinations of agents, including API-based, open-source, and domain-specific models, leading to an 8% improvement on MATH. Finally, we analyze the individual components of ReConcile, demonstrating that the diversity originating from different models is critical to its superior performance. Code: https://github.com/dinobby/ReConcile | http://arxiv.org/abs/2309.13007v2 | "2023-09-22T17:12:45Z" | cs.CL, cs.AI, cs.LG | 2,023 |
Self-Explanation Prompting Improves Dialogue Understanding in Large Language Models | Haoyu Gao, Ting-En Lin, Hangyu Li, Min Yang, Yuchuan Wu, Wentao Ma, Yongbin Li | Task-oriented dialogue (TOD) systems facilitate users in executing various activities via multi-turn dialogues, but Large Language Models (LLMs) often struggle to comprehend these intricate contexts. In this study, we propose a novel "Self-Explanation" prompting strategy to enhance the comprehension abilities of LLMs in multi-turn dialogues. This task-agnostic approach requires the model to analyze each dialogue utterance before task execution, thereby improving performance across various dialogue-centric tasks. Experimental results from six benchmark datasets confirm that our method consistently outperforms other zero-shot prompts and matches or exceeds the efficacy of few-shot prompts, demonstrating its potential as a powerful tool in enhancing LLMs' comprehension in complex dialogue tasks. | http://arxiv.org/abs/2309.12940v1 | "2023-09-22T15:41:34Z" | cs.CL, cs.AI | 2,023 |
TOPFORMER: Topology-Aware Authorship Attribution of Deepfake Texts with Diverse Writing Styles | Adaku Uchendu, Thai Le, Dongwon Lee | Recent advances in Large Language Models (LLMs) have enabled the generation of open-ended high-quality texts, that are non-trivial to distinguish from human-written texts. We refer to such LLM-generated texts as deepfake texts. There are currently over 72K text generation models in the huggingface model repo. As such, users with malicious intent can easily use these open-sourced LLMs to generate harmful texts and dis/misinformation at scale. To mitigate this problem, a computational method to determine if a given text is a deepfake text or not is desired--i.e., Turing Test (TT). In particular, in this work, we investigate the more general version of the problem, known as Authorship Attribution (AA), in a multi-class setting--i.e., not only determining if a given text is a deepfake text or not but also being able to pinpoint which LLM is the author. We propose TopFormer to improve existing AA solutions by capturing more linguistic patterns in deepfake texts by including a Topological Data Analysis (TDA) layer in the Transformer-based model. We show the benefits of having a TDA layer when dealing with imbalanced, and multi-style datasets, by extracting TDA features from the reshaped $pooled\_output$ of our backbone as input. This Transformer-based model captures contextual representations (i.e., semantic and syntactic linguistic features), while TDA captures the shape and structure of data (i.e., linguistic structures). Finally, TopFormer, outperforms all baselines in all 3 datasets, achieving up to 7\% increase in Macro F1 score. | http://arxiv.org/abs/2309.12934v2 | "2023-09-22T15:32:49Z" | cs.CL | 2,023 |
Affect Recognition in Conversations Using Large Language Models | Shutong Feng, Guangzhi Sun, Nurul Lubis, Chao Zhang, Milica Gašić | Affect recognition, encompassing emotions, moods, and feelings, plays a pivotal role in human communication. In the realm of conversational artificial intelligence (AI), the ability to discern and respond to human affective cues is a critical factor for creating engaging and empathetic interactions. This study delves into the capacity of large language models (LLMs) to recognise human affect in conversations, with a focus on both open-domain chit-chat dialogues and task-oriented dialogues. Leveraging three diverse datasets, namely IEMOCAP, EmoWOZ, and DAIC-WOZ, covering a spectrum of dialogues from casual conversations to clinical interviews, we evaluated and compared LLMs' performance in affect recognition. Our investigation explores the zero-shot and few-shot capabilities of LLMs through in-context learning (ICL) as well as their model capacities through task-specific fine-tuning. Additionally, this study takes into account the potential impact of automatic speech recognition (ASR) errors on LLM predictions. With this work, we aim to shed light on the extent to which LLMs can replicate human-like affect recognition capabilities in conversations. | http://arxiv.org/abs/2309.12881v1 | "2023-09-22T14:11:23Z" | cs.CL | 2,023 |
AnglE-optimized Text Embeddings | Xianming Li, Jing Li | High-quality text embedding is pivotal in improving semantic textual similarity (STS) tasks, which are crucial components in Large Language Model (LLM) applications. However, a common challenge existing text embedding models face is the problem of vanishing gradients, primarily due to their reliance on the cosine function in the optimization objective, which has saturation zones. To address this issue, this paper proposes a novel angle-optimized text embedding model called AnglE. The core idea of AnglE is to introduce angle optimization in a complex space. This novel approach effectively mitigates the adverse effects of the saturation zone in the cosine function, which can impede gradient and hinder optimization processes. To set up a comprehensive STS evaluation, we experimented on existing short-text STS datasets and a newly collected long-text STS dataset from GitHub Issues. Furthermore, we examine domain-specific STS scenarios with limited labeled data and explore how AnglE works with LLM-annotated data. Extensive experiments were conducted on various tasks including short-text STS, long-text STS, and domain-specific STS tasks. The results show that AnglE outperforms the state-of-the-art (SOTA) STS models that ignore the cosine saturation zone. These findings demonstrate the ability of AnglE to generate high-quality text embeddings and the usefulness of angle optimization in STS. | http://arxiv.org/abs/2309.12871v6 | "2023-09-22T13:52:42Z" | cs.CL, cs.AI, cs.LG | 2,023 |
Furthest Reasoning with Plan Assessment: Stable Reasoning Path with Retrieval-Augmented Large Language Models | Yin Zhu, Zhiling Luo, Gong Cheng | Large Language Models (LLMs), acting as a powerful reasoner and generator, exhibit extraordinary performance across various natural language tasks, such as question answering (QA). Among these tasks, Multi-Hop Question Answering (MHQA) stands as a widely discussed category, necessitating seamless integration between LLMs and the retrieval of external knowledge. Existing methods employ LLM to generate reasoning paths and plans, and utilize IR to iteratively retrieve related knowledge, but these approaches have inherent flaws. On one hand, Information Retriever (IR) is hindered by the low quality of generated queries by LLM. On the other hand, LLM is easily misguided by the irrelevant knowledge by IR. These inaccuracies, accumulated by the iterative interaction between IR and LLM, lead to a disaster in effectiveness at the end. To overcome above barriers, in this paper, we propose a novel pipeline for MHQA called Furthest-Reasoning-with-Plan-Assessment (FuRePA), including an improved framework (Furthest Reasoning) and an attached module (Plan Assessor). 1) Furthest reasoning operates by masking previous reasoning path and generated queries for LLM, encouraging LLM generating chain of thought from scratch in each iteration. This approach enables LLM to break the shackle built by previous misleading thoughts and queries (if any). 2) The Plan Assessor is a trained evaluator that selects an appropriate plan from a group of candidate plans proposed by LLM. Our methods are evaluated on three highly recognized public multi-hop question answering datasets and outperform state-of-the-art on most metrics (achieving a 10%-12% in answer accuracy). | http://arxiv.org/abs/2309.12767v1 | "2023-09-22T10:15:13Z" | cs.CL | 2,023 |
In-context Interference in Chat-based Large Language Models | Eric Nuertey Coleman, Julio Hurtado, Vincenzo Lomonaco | Large language models (LLMs) have had a huge impact on society due to their impressive capabilities and vast knowledge of the world. Various applications and tools have been created that allow users to interact with these models in a black-box scenario. However, one limitation of this scenario is that users cannot modify the internal knowledge of the model, and the only way to add or modify internal knowledge is by explicitly mentioning it to the model during the current interaction. This learning process is called in-context training, and it refers to training that is confined to the user's current session or context. In-context learning has significant applications, but also has limitations that are seldom studied. In this paper, we present a study that shows how the model can suffer from interference between information that continually flows in the context, causing it to forget previously learned knowledge, which can reduce the model's performance. Along with showing the problem, we propose an evaluation benchmark based on the bAbI dataset. | http://arxiv.org/abs/2309.12727v1 | "2023-09-22T09:18:55Z" | cs.AI, cs.CL | 2,023 |
HRoT: Hybrid prompt strategy and Retrieval of Thought for Table-Text Hybrid Question Answering | Tongxu Luo, Fangyu Lei, Jiahe Lei, Weihao Liu, Shihu He, Jun Zhao, Kang Liu | Answering numerical questions over hybrid contents from the given tables and text(TextTableQA) is a challenging task. Recently, Large Language Models (LLMs) have gained significant attention in the NLP community. With the emergence of large language models, In-Context Learning and Chain-of-Thought prompting have become two particularly popular research topics in this field. In this paper, we introduce a new prompting strategy called Hybrid prompt strategy and Retrieval of Thought for TextTableQA. Through In-Context Learning, we prompt the model to develop the ability of retrieval thinking when dealing with hybrid data. Our method achieves superior performance compared to the fully-supervised SOTA on the MultiHiertt dataset in the few-shot setting. | http://arxiv.org/abs/2309.12669v1 | "2023-09-22T07:26:17Z" | cs.CL | 2,023 |
Can LLMs Augment Low-Resource Reading Comprehension Datasets? Opportunities and Challenges | Vinay Samuel, Houda Aynaou, Arijit Ghosh Chowdhury, Karthik Venkat Ramanan, Aman Chadha | Large Language Models (LLMs) have demonstrated impressive zero shot performance on a wide range of NLP tasks, demonstrating the ability to reason and apply commonsense. A relevant application is to use them for creating high quality synthetic datasets for downstream tasks. In this work, we probe whether GPT-4 can be used to augment existing extractive reading comprehension datasets. Automating data annotation processes has the potential to save large amounts of time, money and effort that goes into manually labelling datasets. In this paper, we evaluate the performance of GPT-4 as a replacement for human annotators for low resource reading comprehension tasks, by comparing performance after fine tuning, and the cost associated with annotation. This work serves to be the first analysis of LLMs as synthetic data augmenters for QA systems, highlighting the unique opportunities and challenges. Additionally, we release augmented versions of low resource datasets, that will allow the research community to create further benchmarks for evaluation of generated datasets. | http://arxiv.org/abs/2309.12426v1 | "2023-09-21T18:48:02Z" | cs.CL, cs.AI | 2,023 |
LongLoRA: Efficient Fine-tuning of Long-Context Large Language Models | Yukang Chen, Shengju Qian, Haotian Tang, Xin Lai, Zhijian Liu, Song Han, Jiaya Jia | We present LongLoRA, an efficient fine-tuning approach that extends the context sizes of pre-trained large language models (LLMs), with limited computation cost. Typically, training LLMs with long context sizes is computationally expensive, requiring extensive training hours and GPU resources. For example, training on the context length of 8192 needs 16x computational costs in self-attention layers as that of 2048. In this paper, we speed up the context extension of LLMs in two aspects. On the one hand, although dense global attention is needed during inference, fine-tuning the model can be effectively and efficiently done by sparse local attention. The proposed shifted sparse attention effectively enables context extension, leading to non-trivial computation saving with similar performance to fine-tuning with vanilla attention. Particularly, it can be implemented with only two lines of code in training, while being optional in inference. On the other hand, we revisit the parameter-efficient fine-tuning regime for context expansion. Notably, we find that LoRA for context extension works well under the premise of trainable embedding and normalization. LongLoRA combines this improved LoRA with S^2-Attn. LongLoRA demonstrates strong empirical results on various tasks on Llama2 models from 7B/13B to 70B. LongLoRA extends Llama2 7B from 4k context to 100k, or Llama2 70B to 32k on a single 8x A100 machine. LongLoRA extends models' context while retaining their original architectures, and is compatible with most existing techniques, like Flash-Attention2. In addition, we further conduct supervised fine-tuning with LongLoRA and our long instruction-following LongAlpaca dataset. | http://arxiv.org/abs/2309.12307v3 | "2023-09-21T17:59:11Z" | cs.CL, cs.AI, cs.LG | 2,023 |
Reranking for Natural Language Generation from Logical Forms: A Study based on Large Language Models | Levon Haroutunian, Zhuang Li, Lucian Galescu, Philip Cohen, Raj Tumuluri, Gholamreza Haffari | Large language models (LLMs) have demonstrated impressive capabilities in natural language generation. However, their output quality can be inconsistent, posing challenges for generating natural language from logical forms (LFs). This task requires the generated outputs to embody the exact semantics of LFs, without missing any LF semantics or creating any hallucinations. In this work, we tackle this issue by proposing a novel generate-and-rerank approach. Our approach involves initially generating a set of candidate outputs by prompting an LLM and subsequently reranking them using a task-specific reranker model. In addition, we curate a manually collected dataset to evaluate the alignment between different ranking metrics and human judgements. The chosen ranking metrics are utilized to enhance the training and evaluation of the reranker model. By conducting extensive experiments on three diverse datasets, we demonstrate that the candidates selected by our reranker outperform those selected by baseline methods in terms of semantic consistency and fluency, as measured by three comprehensive metrics. Our findings provide strong evidence for the effectiveness of our approach in improving the quality of generated outputs. | http://arxiv.org/abs/2309.12294v1 | "2023-09-21T17:54:58Z" | cs.CL | 2,023 |
MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Models | Longhui Yu, Weisen Jiang, Han Shi, Jincheng Yu, Zhengying Liu, Yu Zhang, James T. Kwok, Zhenguo Li, Adrian Weller, Weiyang Liu | Large language models (LLMs) have pushed the limits of natural language understanding and exhibited excellent problem-solving ability. Despite the great success, most existing open-source LLMs (e.g., LLaMA-2) are still far away from satisfactory for solving mathematical problem due to the complex reasoning procedures. To bridge this gap, we propose MetaMath, a fine-tuned language model that specializes in mathematical reasoning. Specifically, we start by bootstrapping mathematical questions by rewriting the question from multiple perspectives without extra knowledge, which results in a new dataset called MetaMathQA. Then we fine-tune the LLaMA-2 models on MetaMathQA. Experimental results on two popular benchmarks (i.e., GSM8K and MATH) for mathematical reasoning demonstrate that MetaMath outperforms a suite of open-source LLMs by a significant margin. Our MetaMath-7B model achieves 66.4% on GSM8K and 19.4% on MATH, exceeding the state-of-the-art models of the same size by 11.5% and 8.7%. Particularly, MetaMath-70B achieves an accuracy of 82.3% on GSM8K, slightly better than GPT-3.5-Turbo. We release all the MetaMathQA dataset, the MetaMath models with different model sizes and the training code for public use. | http://arxiv.org/abs/2309.12284v4 | "2023-09-21T17:45:42Z" | cs.CL, cs.AI | 2,023 |
Bad Actor, Good Advisor: Exploring the Role of Large Language Models in Fake News Detection | Beizhe Hu, Qiang Sheng, Juan Cao, Yuhui Shi, Yang Li, Danding Wang, Peng Qi | Detecting fake news requires both a delicate sense of diverse clues and a profound understanding of the real-world background, which remains challenging for detectors based on small language models (SLMs) due to their knowledge and capability limitations. Recent advances in large language models (LLMs) have shown remarkable performance in various tasks, but whether and how LLMs could help with fake news detection remains underexplored. In this paper, we investigate the potential of LLMs in fake news detection. First, we conduct an empirical study and find that a sophisticated LLM such as GPT 3.5 could generally expose fake news and provide desirable multi-perspective rationales but still underperforms the basic SLM, fine-tuned BERT. Our subsequent analysis attributes such a gap to the LLM's inability to select and integrate rationales properly to conclude. Based on these findings, we propose that current LLMs may not substitute fine-tuned SLMs in fake news detection but can be a good advisor for SLMs by providing multi-perspective instructive rationales. To instantiate this proposal, we design an adaptive rationale guidance network for fake news detection (ARG), in which SLMs selectively acquire insights on news analysis from the LLMs' rationales. We further derive a rationale-free version of ARG by distillation, namely ARG-D, which services cost-sensitive scenarios without querying LLMs. Experiments on two real-world datasets demonstrate that ARG and ARG-D outperform three types of baseline methods, including SLM-based, LLM-based, and combinations of small and large language models. | http://arxiv.org/abs/2309.12247v2 | "2023-09-21T16:47:30Z" | cs.CL, cs.AI, cs.CY | 2,023 |
Code Soliloquies for Accurate Calculations in Large Language Models | Shashank Sonkar, MyCo Le, Xinghe Chen, Naiming Liu, Debshila Basu Mallick, Richard G. Baraniuk | High-quality conversational datasets are crucial for the successful development of Intelligent Tutoring Systems (ITS) that utilize a Large Language Model (LLM) backend. Synthetic student-teacher dialogues, generated using advanced GPT-4 models, are a common strategy for creating these datasets. However, subjects like physics that entail complex calculations pose a challenge. While GPT-4 presents impressive language processing capabilities, its limitations in fundamental mathematical reasoning curtail its efficacy for such subjects. To tackle this limitation, we introduce in this paper an innovative stateful prompt design. Our design orchestrates a mock conversation where both student and tutorbot roles are simulated by GPT-4. Each student response triggers an internal monologue, or `code soliloquy' in the GPT-tutorbot, which assesses whether its subsequent response would necessitate calculations. If a calculation is deemed necessary, it scripts the relevant Python code and uses the Python output to construct a response to the student. Our approach notably enhances the quality of synthetic conversation datasets, especially for subjects that are calculation-intensive. Our preliminary Subject Matter Expert evaluations reveal that our Higgs model, a fine-tuned LLaMA model, effectively uses Python for computations, which significantly enhances the accuracy and computational reliability of Higgs' responses. Code, models, and datasets is available at https://github.com/luffycodes/Tutorbot-Spock-Phys. | http://arxiv.org/abs/2309.12161v2 | "2023-09-21T15:16:58Z" | cs.CL | 2,023 |
PEFTT: Parameter-Efficient Fine-Tuning for low-resource Tibetan pre-trained language models | Zhou Mingjun, Daiqing Zhuoma, Qun Nuo, Nyima Tashi | In this era of large language models (LLMs), the traditional training of models has become increasingly unimaginable for regular users and institutions. The exploration of efficient fine-tuning for high-resource languages on these models is an undeniable trend that is gradually gaining popularity. However, there has been very little exploration for various low-resource languages, such as Tibetan. Research in Tibetan NLP is inherently scarce and limited. While there is currently no existing large language model for Tibetan due to its low-resource nature, that day will undoubtedly arrive. Therefore, research on efficient fine-tuning for low-resource language models like Tibetan is highly necessary. Our research can serve as a reference to fill this crucial gap. Efficient fine-tuning strategies for pre-trained language models (PLMs) in Tibetan have seen minimal exploration. We conducted three types of efficient fine-tuning experiments on the publicly available TNCC-title dataset: "prompt-tuning," "Adapter lightweight fine-tuning," and "prompt-tuning + Adapter fine-tuning." The experimental results demonstrate significant improvements using these methods, providing valuable insights for advancing Tibetan language applications in the context of pre-trained models. | http://arxiv.org/abs/2309.12109v1 | "2023-09-21T14:29:23Z" | cs.CL, cs.AI | 2,023 |
Prompt Tuned Embedding Classification for Multi-Label Industry Sector Allocation | Valentin Leonhard Buchner, Lele Cao, Jan-Christoph Kalo, Vilhelm von Ehrenheim | Prompt Tuning is emerging as a scalable and cost-effective method to fine-tune Pretrained Language Models (PLMs), which are often referred to as Large Language Models (LLMs). This study benchmarks the performance and computational efficiency of Prompt Tuning and baselines for multi-label text classification. This is applied to the challenging task of classifying companies into an investment firm's proprietary industry taxonomy, supporting their thematic investment strategy. Text-to-text classification is frequently reported to outperform task-specific classification heads, but has several limitations when applied to a multi-label classification problem where each label consists of multiple tokens: (a) Generated labels may not match any label in the label taxonomy; (b) The fine-tuning process lacks permutation invariance and is sensitive to the order of the provided labels; (c) The model provides binary decisions rather than appropriate confidence scores. Limitation (a) is addressed by applying constrained decoding using Trie Search, which slightly improves classification performance. All limitations (a), (b), and (c) are addressed by replacing the PLM's language head with a classification head, which is referred to as Prompt Tuned Embedding Classification (PTEC). This improves performance significantly, while also reducing computational costs during inference. In our industrial application, the training data is skewed towards well-known companies. We confirm that the model's performance is consistent across both well-known and less-known companies. Our overall results indicate the continuing need to adapt state-of-the-art methods to domain-specific tasks, even in the era of PLMs with strong generalization abilities. We release our codebase and a benchmarking dataset at https://github.com/EQTPartners/PTEC. | http://arxiv.org/abs/2309.12075v3 | "2023-09-21T13:45:32Z" | cs.CL, cs.AI, 68T50, I.2.7; I.2.0 | 2,023 |
LMSYS-Chat-1M: A Large-Scale Real-World LLM Conversation Dataset | Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Tianle Li, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zhuohan Li, Zi Lin, Eric P. Xing, Joseph E. Gonzalez, Ion Stoica, Hao Zhang | Studying how people interact with large language models (LLMs) in real-world scenarios is increasingly important due to their widespread use in various applications. In this paper, we introduce LMSYS-Chat-1M, a large-scale dataset containing one million real-world conversations with 25 state-of-the-art LLMs. This dataset is collected from 210K unique IP addresses in the wild on our Vicuna demo and Chatbot Arena website. We offer an overview of the dataset's content, including its curation process, basic statistics, and topic distribution, highlighting its diversity, originality, and scale. We demonstrate its versatility through four use cases: developing content moderation models that perform similarly to GPT-4, building a safety benchmark, training instruction-following models that perform similarly to Vicuna, and creating challenging benchmark questions. We believe that this dataset will serve as a valuable resource for understanding and advancing LLM capabilities. The dataset is publicly available at https://huggingface.co/datasets/lmsys/lmsys-chat-1m. | http://arxiv.org/abs/2309.11998v4 | "2023-09-21T12:13:55Z" | cs.CL, cs.AI | 2,023 |
InstructERC: Reforming Emotion Recognition in Conversation with a Retrieval Multi-task LLMs Framework | Shanglin Lei, Guanting Dong, Xiaoping Wang, Keheng Wang, Sirui Wang | The field of emotion recognition of conversation (ERC) has been focusing on separating sentence feature encoding and context modeling, lacking exploration in generative paradigms based on unified designs. In this study, we propose a novel approach, \textbf{InstructERC}, to reformulate the ERC task from a discriminative framework to a generative framework based on Large Language Models (LLMs). InstructERC makes three significant contributions: (1) it introduces a simple yet effective retrieval template module, which helps the model explicitly integrate multi-granularity dialogue supervision information. (2) We introduce two additional emotion alignment tasks, namely speaker identification and emotion prediction tasks, to implicitly model the dialogue role relationships and future emotional tendencies in conversations. (3) Pioneeringly, we unify emotion labels across benchmarks through the feeling wheel to fit real application scenarios. InstructERC still perform impressively on this unified dataset. Our LLM-based plugin framework significantly outperforms all previous models and achieves comprehensive SOTA on three commonly used ERC datasets. Extensive analysis of parameter-efficient and data-scaling experiments provides empirical guidance for applying it in practical scenarios. Our code and aligned unified dataset (UIME) can be found in the Github link.\footnote{You can find the offical realization in the Github link: https://github.com/LIN-SHANG/InstructERC} | http://arxiv.org/abs/2309.11911v4 | "2023-09-21T09:22:07Z" | cs.CL | 2,023 |
Goal-Oriented Prompt Attack and Safety Evaluation for LLMs | Chengyuan Liu, Fubang Zhao, Lizhi Qing, Yangyang Kang, Changlong Sun, Kun Kuang, Fei Wu | Large Language Models (LLMs) presents significant priority in text understanding and generation. However, LLMs suffer from the risk of generating harmful contents especially while being employed to applications. There are several black-box attack methods, such as Prompt Attack, which can change the behaviour of LLMs and induce LLMs to generate unexpected answers with harmful contents. Researchers are interested in Prompt Attack and Defense with LLMs, while there is no publicly available dataset with high successful attacking rate to evaluate the abilities of defending prompt attack. In this paper, we introduce a pipeline to construct high-quality prompt attack samples, along with a Chinese prompt attack dataset called CPAD. Our prompts aim to induce LLMs to generate unexpected outputs with several carefully designed prompt attack templates and widely concerned attacking contents. Different from previous datasets involving safety estimation, we construct the prompts considering three dimensions: contents, attacking methods and goals. Especially, the attacking goals indicate the behaviour expected after successfully attacking the LLMs, thus the responses can be easily evaluated and analysed. We run several popular Chinese LLMs on our dataset, and the results show that our prompts are significantly harmful to LLMs, with around 70% attack success rate to GPT-3.5. CPAD is publicly available at https://github.com/liuchengyuan123/CPAD. | http://arxiv.org/abs/2309.11830v2 | "2023-09-21T07:07:49Z" | cs.CL | 2,023 |
Privacy-Preserving In-Context Learning with Differentially Private Few-Shot Generation | Xinyu Tang, Richard Shin, Huseyin A. Inan, Andre Manoel, Fatemehsadat Mireshghallah, Zinan Lin, Sivakanth Gopi, Janardhan Kulkarni, Robert Sim | We study the problem of in-context learning (ICL) with large language models (LLMs) on private datasets. This scenario poses privacy risks, as LLMs may leak or regurgitate the private examples demonstrated in the prompt. We propose a novel algorithm that generates synthetic few-shot demonstrations from the private dataset with formal differential privacy (DP) guarantees, and show empirically that it can achieve effective ICL. We conduct extensive experiments on standard benchmarks and compare our algorithm with non-private ICL and zero-shot solutions. Our results demonstrate that our algorithm can achieve competitive performance with strong privacy levels. These results open up new possibilities for ICL with privacy protection for a broad range of applications. | http://arxiv.org/abs/2309.11765v2 | "2023-09-21T03:59:00Z" | cs.LG, cs.CR | 2,023 |
Choice-75: A Dataset on Decision Branching in Script Learning | Zhaoyi Joey Hou, Li Zhang, Chris Callison-Burch | Script learning studies how stereotypical events unfold, enabling machines to reason about narratives with implicit information. Previous works mostly consider a script as a linear sequence of events while ignoring the potential branches that arise due to people's circumstantial choices. We hence propose Choice-75, the first benchmark that challenges intelligent systems to make decisions given descriptive scenarios, containing 75 scripts and more than 600 scenarios. We also present preliminary results with current large language models (LLM). Although they demonstrate overall decent performance, there is still notable headroom in hard scenarios. | http://arxiv.org/abs/2309.11737v2 | "2023-09-21T02:23:44Z" | cs.AI | 2,023 |
A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models | Haoran Xu, Young Jin Kim, Amr Sharaf, Hany Hassan Awadalla | Generative Large Language Models (LLMs) have achieved remarkable advancements in various NLP tasks. However, these advances have not been reflected in the translation task, especially those with moderate model sizes (i.e., 7B or 13B parameters), which still lag behind conventional supervised encoder-decoder translation models. Previous studies have attempted to improve the translation capabilities of these moderate LLMs, but their gains have been limited. In this study, we propose a novel fine-tuning approach for LLMs that is specifically designed for the translation task, eliminating the need for the abundant parallel data that traditional translation models usually depend on. Our approach consists of two fine-tuning stages: initial fine-tuning on monolingual data followed by subsequent fine-tuning on a small set of high-quality parallel data. We introduce the LLM developed through this strategy as Advanced Language Model-based trAnslator (ALMA). Based on LLaMA-2 as our underlying model, our results show that the model can achieve an average improvement of more than 12 BLEU and 12 COMET over its zero-shot performance across 10 translation directions from the WMT'21 (2 directions) and WMT'22 (8 directions) test datasets. The performance is significantly better than all prior work and even superior to the NLLB-54B model and GPT-3.5-text-davinci-003, with only 7B or 13B parameters. This method establishes the foundation for a novel training paradigm in machine translation. | http://arxiv.org/abs/2309.11674v2 | "2023-09-20T22:53:15Z" | cs.CL | 2,023 |
Construction of Paired Knowledge Graph-Text Datasets Informed by Cyclic Evaluation | Ali Mousavi, Xin Zhan, He Bai, Peng Shi, Theo Rekatsinas, Benjamin Han, Yunyao Li, Jeff Pound, Josh Susskind, Natalie Schluter, Ihab Ilyas, Navdeep Jaitly | Datasets that pair Knowledge Graphs (KG) and text together (KG-T) can be used to train forward and reverse neural models that generate text from KG and vice versa. However models trained on datasets where KG and text pairs are not equivalent can suffer from more hallucination and poorer recall. In this paper, we verify this empirically by generating datasets with different levels of noise and find that noisier datasets do indeed lead to more hallucination. We argue that the ability of forward and reverse models trained on a dataset to cyclically regenerate source KG or text is a proxy for the equivalence between the KG and the text in the dataset. Using cyclic evaluation we find that manually created WebNLG is much better than automatically created TeKGen and T-REx. Guided by these observations, we construct a new, improved dataset called LAGRANGE using heuristics meant to improve equivalence between KG and text and show the impact of each of the heuristics on cyclic evaluation. We also construct two synthetic datasets using large language models (LLMs), and observe that these are conducive to models that perform significantly well on cyclic generation of text, but less so on cyclic generation of KGs, probably because of a lack of a consistent underlying ontology. | http://arxiv.org/abs/2309.11669v1 | "2023-09-20T22:30:20Z" | cs.CL | 2,023 |
Towards Effective Disambiguation for Machine Translation with Large Language Models | Vivek Iyer, Pinzhen Chen, Alexandra Birch | Resolving semantic ambiguity has long been recognised as a central challenge in the field of Machine Translation. Recent work on benchmarking translation performance on ambiguous sentences has exposed the limitations of conventional Neural Machine Translation (NMT) systems, which fail to handle many such cases. Large language models (LLMs) have emerged as a promising alternative, demonstrating comparable performance to traditional NMT models while introducing new paradigms for controlling the target outputs. In this paper, we study the capabilities of LLMs to translate "ambiguous sentences" - i.e. those containing highly polysemous words and/or rare word senses. We also propose two ways to improve their disambiguation capabilities, through a) in-context learning and b) fine-tuning on carefully curated ambiguous datasets. Experiments show that our methods can match or outperform state-of-the-art systems such as DeepL and NLLB in four out of five language directions. Our research provides valuable insights into effectively adapting LLMs to become better disambiguators during Machine Translation. We release our curated disambiguation corpora and resources at https://data.statmt.org/ambiguous-europarl. | http://arxiv.org/abs/2309.11668v2 | "2023-09-20T22:22:52Z" | cs.CL | 2,023 |
Controlled Generation with Prompt Insertion for Natural Language Explanations in Grammatical Error Correction | Masahiro Kaneko, Naoaki Okazaki | In Grammatical Error Correction (GEC), it is crucial to ensure the user's comprehension of a reason for correction. Existing studies present tokens, examples, and hints as to the basis for correction but do not directly explain the reasons for corrections. Although methods that use Large Language Models (LLMs) to provide direct explanations in natural language have been proposed for various tasks, no such method exists for GEC. Generating explanations for GEC corrections involves aligning input and output tokens, identifying correction points, and presenting corresponding explanations consistently. However, it is not straightforward to specify a complex format to generate explanations, because explicit control of generation is difficult with prompts. This study introduces a method called controlled generation with Prompt Insertion (PI) so that LLMs can explain the reasons for corrections in natural language. In PI, LLMs first correct the input text, and then we automatically extract the correction points based on the rules. The extracted correction points are sequentially inserted into the LLM's explanation output as prompts, guiding the LLMs to generate explanations for the correction points. We also create an Explainable GEC (XGEC) dataset of correction reasons by annotating NUCLE, CoNLL2013, and CoNLL2014. Although generations from GPT-3 and ChatGPT using original prompts miss some correction points, the generation control using PI can explicitly guide to describe explanations for all correction points, contributing to improved performance in generating correction reasons. | http://arxiv.org/abs/2309.11439v1 | "2023-09-20T16:14:10Z" | cs.CL | 2,023 |
DISC-LawLLM: Fine-tuning Large Language Models for Intelligent Legal Services | Shengbin Yue, Wei Chen, Siyuan Wang, Bingxuan Li, Chenchen Shen, Shujun Liu, Yuxuan Zhou, Yao Xiao, Song Yun, Xuanjing Huang, Zhongyu Wei | We propose DISC-LawLLM, an intelligent legal system utilizing large language models (LLMs) to provide a wide range of legal services. We adopt legal syllogism prompting strategies to construct supervised fine-tuning datasets in the Chinese Judicial domain and fine-tune LLMs with legal reasoning capability. We augment LLMs with a retrieval module to enhance models' ability to access and utilize external legal knowledge. A comprehensive legal benchmark, DISC-Law-Eval, is presented to evaluate intelligent legal systems from both objective and subjective dimensions. Quantitative and qualitative results on DISC-Law-Eval demonstrate the effectiveness of our system in serving various users across diverse legal scenarios. The detailed resources are available at https://github.com/FudanDISC/DISC-LawLLM. | http://arxiv.org/abs/2309.11325v2 | "2023-09-20T13:50:26Z" | cs.CL | 2,023 |
StructChart: Perception, Structuring, Reasoning for Visual Chart Understanding | Renqiu Xia, Bo Zhang, Haoyang Peng, Hancheng Ye, Xiangchao Yan, Peng Ye, Botian Shi, Yu Qiao, Junchi Yan | Charts are common in literature across different scientific fields, conveying rich information easily accessible to readers. Current chart-related tasks focus on either chart perception which refers to extracting information from the visual charts, or performing reasoning given the extracted data, e.g. in a tabular form. In this paper, we aim to establish a unified and label-efficient learning paradigm for joint perception and reasoning tasks, which can be generally applicable to different downstream tasks, beyond the question-answering task as specifically studied in peer works. Specifically, StructChart first reformulates the chart information from the popular tubular form (specifically linearized CSV) to the proposed Structured Triplet Representations (STR), which is more friendly for reducing the task gap between chart perception and reasoning due to the employed structured information extraction for charts. We then propose a Structuring Chart-oriented Representation Metric (SCRM) to quantitatively evaluate the performance for the chart perception task. To enrich the dataset for training, we further explore the possibility of leveraging the Large Language Model (LLM), enhancing the chart diversity in terms of both chart visual style and its statistical information. Extensive experiments are conducted on various chart-related tasks, demonstrating the effectiveness and promising potential for a unified chart perception-reasoning paradigm to push the frontier of chart understanding. | http://arxiv.org/abs/2309.11268v4 | "2023-09-20T12:51:13Z" | cs.CV | 2,023 |
Are Large Language Models Really Robust to Word-Level Perturbations? | Haoyu Wang, Guozheng Ma, Cong Yu, Ning Gui, Linrui Zhang, Zhiqi Huang, Suwei Ma, Yongzhe Chang, Sen Zhang, Li Shen, Xueqian Wang, Peilin Zhao, Dacheng Tao | The swift advancement in the scales and capabilities of Large Language Models (LLMs) positions them as promising tools for a variety of downstream tasks. In addition to the pursuit of better performance and the avoidance of violent feedback on a certain prompt, to ensure the responsibility of the LLM, much attention is drawn to the robustness of LLMs. However, existing evaluation methods mostly rely on traditional question answering datasets with predefined supervised labels, which do not align with the superior generation capabilities of contemporary LLMs. To address this issue, we propose a novel rational evaluation approach that leverages pre-trained reward models as diagnostic tools to evaluate the longer conversation generated from more challenging open questions by LLMs, which we refer to as the Reward Model for Reasonable Robustness Evaluation (TREvaL). Longer conversations manifest the comprehensive grasp of language models in terms of their proficiency in understanding questions, a capability not entirely encompassed by individual words or letters, which may exhibit oversimplification and inherent biases. Our extensive empirical experiments demonstrate that TREvaL provides an innovative method for evaluating the robustness of an LLM. Furthermore, our results demonstrate that LLMs frequently exhibit vulnerability to word-level perturbations that are commonplace in daily language usage. Notably, we are surprised to discover that robustness tends to decrease as fine-tuning (SFT and RLHF) is conducted. The code of TREval is available in https://github.com/Harry-mic/TREvaL. | http://arxiv.org/abs/2309.11166v2 | "2023-09-20T09:23:46Z" | cs.CL, cs.AI | 2,023 |
XATU: A Fine-grained Instruction-based Benchmark for Explainable Text Updates | Haopeng Zhang, Hayate Iso, Sairam Gurajada, Nikita Bhutani | Text editing is a crucial task of modifying text to better align with user intents. However, existing text editing benchmark datasets contain only coarse-grained instructions and lack explainability, thus resulting in outputs that deviate from the intended changes outlined in the gold reference. To comprehensively investigate the text editing capabilities of large language models (LLMs), this paper introduces XATU, the first benchmark specifically designed for fine-grained instruction-based explainable text editing. XATU considers finer-grained text editing tasks of varying difficulty (simplification, grammar check, fact-check, etc.), incorporating lexical, syntactic, semantic, and knowledge-intensive edit aspects. To enhance interpretability, we combine LLM-based annotation and human annotation, resulting in a benchmark that includes fine-grained instructions and gold-standard edit explanations. By evaluating existing LLMs against our benchmark, we demonstrate the effectiveness of instruction tuning and the impact of underlying architecture across various editing tasks. Furthermore, extensive experimentation reveals the significant role of explanations in fine-tuning language models for text editing tasks. The benchmark will be open-sourced to support reproduction and facilitate future research at~\url{https://github.com/megagonlabs/xatu}. | http://arxiv.org/abs/2309.11063v2 | "2023-09-20T04:58:59Z" | cs.CL | 2,023 |
In-Context Learning for Text Classification with Many Labels | Aristides Milios, Siva Reddy, Dzmitry Bahdanau | In-context learning (ICL) using large language models for tasks with many labels is challenging due to the limited context window, which makes it difficult to fit a sufficient number of examples in the prompt. In this paper, we use a pre-trained dense retrieval model to bypass this limitation, giving the model only a partial view of the full label space for each inference call. Testing with recent open-source LLMs (OPT, LLaMA), we set new state of the art performance in few-shot settings for three common intent classification datasets, with no finetuning. We also surpass fine-tuned performance on fine-grained sentiment classification in certain cases. We analyze the performance across number of in-context examples and different model scales, showing that larger models are necessary to effectively and consistently make use of larger context lengths for ICL. By running several ablations, we analyze the model's use of: a) the similarity of the in-context examples to the current input, b) the semantic content of the class names, and c) the correct correspondence between examples and labels. We demonstrate that all three are needed to varying degrees depending on the domain, contrary to certain recent works. | http://arxiv.org/abs/2309.10954v2 | "2023-09-19T22:41:44Z" | cs.CL, cs.LG | 2,023 |
Specializing Small Language Models towards Complex Style Transfer via Latent Attribute Pre-Training | Ruiqi Xu, Yongfeng Huang, Xin Chen, Lin Zhang | In this work, we introduce the concept of complex text style transfer tasks, and constructed complex text datasets based on two widely applicable scenarios. Our dataset is the first large-scale data set of its kind, with 700 rephrased sentences and 1,000 sentences from the game Genshin Impact. While large language models (LLM) have shown promise in complex text style transfer, they have drawbacks such as data privacy concerns, network instability, and high deployment costs. To address these issues, we explore the effectiveness of small models (less than T5-3B) with implicit style pre-training through contrastive learning. We also propose a method for automated evaluation of text generation quality based on alignment with human evaluations using ChatGPT. Finally, we compare our approach with existing methods and show that our model achieves state-of-art performances of few-shot text style transfer models. | http://arxiv.org/abs/2309.10929v1 | "2023-09-19T21:01:40Z" | cs.CL | 2,023 |
End-to-End Speech Recognition Contextualization with Large Language Models | Egor Lakomkin, Chunyang Wu, Yassir Fathullah, Ozlem Kalinli, Michael L. Seltzer, Christian Fuegen | In recent years, Large Language Models (LLMs) have garnered significant attention from the research community due to their exceptional performance and generalization capabilities. In this paper, we introduce a novel method for contextualizing speech recognition models incorporating LLMs. Our approach casts speech recognition as a mixed-modal language modeling task based on a pretrained LLM. We provide audio features, along with optional text tokens for context, to train the system to complete transcriptions in a decoder-only fashion. As a result, the system is implicitly incentivized to learn how to leverage unstructured contextual information during training. Our empirical results demonstrate a significant improvement in performance, with a 6% WER reduction when additional textual context is provided. Moreover, we find that our method performs competitively and improve by 7.5% WER overall and 17% WER on rare words against a baseline contextualized RNN-T system that has been trained on more than twenty five times larger speech dataset. Overall, we demonstrate that by only adding a handful number of trainable parameters via adapters, we can unlock contextualized speech recognition capability for the pretrained LLM while keeping the same text-only input functionality. | http://arxiv.org/abs/2309.10917v1 | "2023-09-19T20:28:57Z" | eess.AS, cs.AI, cs.CL, cs.LG, cs.SD | 2,023 |
SlimPajama-DC: Understanding Data Combinations for LLM Training | Zhiqiang Shen, Tianhua Tao, Liqun Ma, Willie Neiswanger, Zhengzhong Liu, Hongyi Wang, Bowen Tan, Joel Hestness, Natalia Vassilieva, Daria Soboleva, Eric Xing | This paper aims to understand the impacts of various data combinations (e.g., web text, wikipedia, github, books) on the training of large language models using SlimPajama. SlimPajama is a rigorously deduplicated, multi-source dataset, which has been refined and further deduplicated to 627B tokens from the extensive 1.2T tokens RedPajama dataset contributed by Together. We've termed our research as SlimPajama-DC, an empirical analysis designed to uncover fundamental characteristics and best practices associated with employing SlimPajama in the training of large language models. During our research with SlimPajama, two pivotal observations emerged: (1) Global deduplication vs. local deduplication. We analyze and discuss how global (across different sources of datasets) and local (within the single source of dataset) deduplications affect the performance of trained models. (2) Proportions of high-quality/highly-deduplicated multi-source datasets in the combination. To study this, we construct six configurations of SlimPajama dataset and train individual ones using 1.3B Cerebras-GPT model with Alibi and SwiGLU. Our best configuration outperforms the 1.3B model trained on RedPajama using the same number of training tokens by a significant margin. All our 1.3B models are trained on Cerebras 16$\times$ CS-2 cluster with a total of 80 PFLOP/s in bf16 mixed precision. We further extend our discoveries (such as increasing data diversity is crucial after global deduplication) on a 7B model with large batch-size training. Our models and the separate SlimPajama-DC datasets are available at: https://huggingface.co/MBZUAI-LLM and https://huggingface.co/datasets/cerebras/SlimPajama-627B. | http://arxiv.org/abs/2309.10818v2 | "2023-09-19T17:59:54Z" | cs.CL, cs.AI | 2,023 |
Language as the Medium: Multimodal Video Classification through text only | Laura Hanu, Anita L. Verő, James Thewlis | Despite an exciting new wave of multimodal machine learning models, current approaches still struggle to interpret the complex contextual relationships between the different modalities present in videos. Going beyond existing methods that emphasize simple activities or objects, we propose a new model-agnostic approach for generating detailed textual descriptions that captures multimodal video information. Our method leverages the extensive knowledge learnt by large language models, such as GPT-3.5 or Llama2, to reason about textual descriptions of the visual and aural modalities, obtained from BLIP-2, Whisper and ImageBind. Without needing additional finetuning of video-text models or datasets, we demonstrate that available LLMs have the ability to use these multimodal textual descriptions as proxies for ``sight'' or ``hearing'' and perform zero-shot multimodal classification of videos in-context. Our evaluations on popular action recognition benchmarks, such as UCF-101 or Kinetics, show these context-rich descriptions can be successfully used in video understanding tasks. This method points towards a promising new research direction in multimodal classification, demonstrating how an interplay between textual, visual and auditory machine learning models can enable more holistic video understanding. | http://arxiv.org/abs/2309.10783v1 | "2023-09-19T17:32:21Z" | cs.CV, cs.AI, cs.CL | 2,023 |
FRASIMED: a Clinical French Annotated Resource Produced through Crosslingual BERT-Based Annotation Projection | Jamil Zaghir, Mina Bjelogrlic, Jean-Philippe Goldman, Soukaïna Aananou, Christophe Gaudet-Blavignac, Christian Lovis | Natural language processing (NLP) applications such as named entity recognition (NER) for low-resource corpora do not benefit from recent advances in the development of large language models (LLMs) where there is still a need for larger annotated datasets. This research article introduces a methodology for generating translated versions of annotated datasets through crosslingual annotation projection. Leveraging a language agnostic BERT-based approach, it is an efficient solution to increase low-resource corpora with few human efforts and by only using already available open data resources. Quantitative and qualitative evaluations are often lacking when it comes to evaluating the quality and effectiveness of semi-automatic data generation strategies. The evaluation of our crosslingual annotation projection approach showed both effectiveness and high accuracy in the resulting dataset. As a practical application of this methodology, we present the creation of French Annotated Resource with Semantic Information for Medical Entities Detection (FRASIMED), an annotated corpus comprising 2'051 synthetic clinical cases in French. The corpus is now available for researchers and practitioners to develop and refine French natural language processing (NLP) applications in the clinical field (https://zenodo.org/record/8355629), making it the largest open annotated corpus with linked medical concepts in French. | http://arxiv.org/abs/2309.10770v1 | "2023-09-19T17:17:28Z" | cs.CL, cs.AI | 2,023 |
MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language Feedback | Xingyao Wang, Zihan Wang, Jiateng Liu, Yangyi Chen, Lifan Yuan, Hao Peng, Heng Ji | To solve complex tasks, large language models (LLMs) often require multiple rounds of interactions with the user, sometimes assisted by external tools. However, current evaluation protocols often emphasize benchmark performance with single-turn exchanges, neglecting the nuanced interactions among the user, LLMs, and external tools, while also underestimating the importance of natural language feedback from users. These oversights contribute to discrepancies between research benchmark evaluations and real-world use cases. We introduce MINT, a benchmark that evaluates LLMs' ability to solve tasks with multi-turn interactions by (1) using tools and (2) leveraging natural language feedback. To ensure reproducibility, we provide an evaluation framework where LLMs can access tools by executing Python code and receive users' natural language feedback simulated by GPT-4. We repurpose a diverse set of established evaluation datasets focusing on reasoning, coding, and decision-making and carefully curate them into a compact subset for efficient evaluation. Our analysis of 20 open- and closed-source LLMs offers intriguing findings. (a) LLMs generally benefit from tools and language feedback, with performance gains (absolute, same below) of 1-8% for each turn of tool use and 2-17% with natural language feedback. (b) Better single-turn performance does not guarantee better multi-turn performance. (c) Surprisingly, on the LLMs evaluated, supervised instruction-finetuning (SIFT) and reinforcement learning from human feedback (RLHF) generally hurt multi-turn capabilities. We expect MINT can help measure progress and incentivize research in improving LLMs' capabilities in multi-turn interactions, especially for open-source communities where multi-turn human evaluation can be less accessible compared to commercial LLMs with a larger user base. | http://arxiv.org/abs/2309.10691v3 | "2023-09-19T15:25:42Z" | cs.CL, cs.AI, cs.LG | 2,023 |
CFGPT: Chinese Financial Assistant with Large Language Model | Jiangtong Li, Yuxuan Bian, Guoxuan Wang, Yang Lei, Dawei Cheng, Zhijun Ding, Changjun Jiang | Large language models (LLMs) have demonstrated great potential in natural language processing tasks within the financial domain. In this work, we present a Chinese Financial Generative Pre-trained Transformer framework, named CFGPT, which includes a dataset~(CFData) for pre-training and supervised fine-tuning, a financial LLM~(CFLLM) to adeptly manage financial texts, and a deployment framework~(CFAPP) designed to navigate real-world financial applications. The CFData comprising both a pre-training dataset and a supervised fine-tuning dataset, where the pre-training dataset collates Chinese financial data and analytics, alongside a smaller subset of general-purpose text with 584M documents and 141B tokens in total, and the supervised fine-tuning dataset is tailored for six distinct financial tasks, embodying various facets of financial analysis and decision-making with 1.5M instruction pairs and 1.5B tokens in total. The CFLLM, which is based on InternLM-7B to balance the model capability and size, is trained on CFData in two stage, continued pre-training and supervised fine-tuning. The CFAPP is centered on large language models (LLMs) and augmented with additional modules to ensure multifaceted functionality in real-world application. Our codes are released at https://github.com/TongjiFinLab/CFGPT. | http://arxiv.org/abs/2309.10654v2 | "2023-09-19T14:34:01Z" | cs.CL, cs.AI, cs.CE | 2,023 |
A Hierarchical Neural Framework for Classification and its Explanation in Large Unstructured Legal Documents | Nishchal Prasad, Mohand Boughanem, Taoufik Dkaki | Automatic legal judgment prediction and its explanation suffer from the problem of long case documents exceeding tens of thousands of words, in general, and having a non-uniform structure. Predicting judgments from such documents and extracting their explanation becomes a challenging task, more so on documents with no structural annotation. We define this problem as "scarce annotated legal documents" and explore their lack of structural information and their long lengths with a deep-learning-based classification framework which we call MESc; "Multi-stage Encoder-based Supervised with-clustering"; for judgment prediction. We explore the adaptability of LLMs with multi-billion parameters (GPT-Neo, and GPT-J) to legal texts and their intra-domain(legal) transfer learning capacity. Alongside this, we compare their performance and adaptability with MESc and the impact of combining embeddings from their last layers. For such hierarchical models, we also propose an explanation extraction algorithm named ORSE; Occlusion sensitivity-based Relevant Sentence Extractor; based on the input-occlusion sensitivity of the model, to explain the predictions with the most relevant sentences from the document. We explore these methods and test their effectiveness with extensive experiments and ablation studies on legal documents from India, the European Union, and the United States with the ILDC dataset and a subset of the LexGLUE dataset. MESc achieves a minimum total performance gain of approximately 2 points over previous state-of-the-art proposed methods, while ORSE applied on MESc achieves a total average gain of 50% over the baseline explainability scores. | http://arxiv.org/abs/2309.10563v2 | "2023-09-19T12:18:28Z" | cs.IR, cs.LG | 2,023 |
Exploring Iterative Enhancement for Improving Learnersourced Multiple-Choice Question Explanations with Large Language Models | Qiming Bao, Juho Leinonen, Alex Yuxuan Peng, Wanjun Zhong, Gaël Gendron, Timothy Pistotti, Alice Huang, Paul Denny, Michael Witbrock, Jiamou Liu | Large language models exhibit superior capabilities in processing and understanding language, yet their applications in educational contexts remain underexplored. Learnersourcing enhances learning by engaging students in creating their own educational content. When learnersourcing multiple-choice questions, creating explanations for the solution of a question is a crucial step; it helps other students understand the solution and promotes a deeper understanding of related concepts. However, it is often difficult for students to craft effective solution explanations, due to limited subject understanding. To help scaffold the task of automated explanation generation, we present and evaluate a framework called "ILearner-LLM", that iteratively enhances the generated explanations for the given questions with large language models. Comprising an explanation generation model and an explanation evaluation model, the framework generates high-quality student-aligned explanations by iteratively feeding the quality rating score from the evaluation model back into the instruction prompt of the explanation generation model. Experimental results demonstrate the effectiveness of our ILearner-LLM on LLaMA2-13B and GPT-4 to generate higher quality explanations that are closer to those written by students on five PeerWise datasets. Our findings represent a promising path to enrich the learnersourcing experience for students and to enhance the capabilities of large language models for educational applications. | http://arxiv.org/abs/2309.10444v4 | "2023-09-19T09:04:15Z" | cs.AI, cs.CL | 2,023 |
Prompt, Condition, and Generate: Classification of Unsupported Claims with In-Context Learning | Peter Ebert Christensen, Srishti Yadav, Serge Belongie | Unsupported and unfalsifiable claims we encounter in our daily lives can influence our view of the world. Characterizing, summarizing, and -- more generally -- making sense of such claims, however, can be challenging. In this work, we focus on fine-grained debate topics and formulate a new task of distilling, from such claims, a countable set of narratives. We present a crowdsourced dataset of 12 controversial topics, comprising more than 120k arguments, claims, and comments from heterogeneous sources, each annotated with a narrative label. We further investigate how large language models (LLMs) can be used to synthesise claims using In-Context Learning. We find that generated claims with supported evidence can be used to improve the performance of narrative classification models and, additionally, that the same model can infer the stance and aspect using a few training examples. Such a model can be useful in applications which rely on narratives , e.g. fact-checking. | http://arxiv.org/abs/2309.10359v1 | "2023-09-19T06:42:37Z" | cs.CL | 2,023 |
Investigating the Catastrophic Forgetting in Multimodal Large Language Models | Yuexiang Zhai, Shengbang Tong, Xiao Li, Mu Cai, Qing Qu, Yong Jae Lee, Yi Ma | Following the success of GPT4, there has been a surge in interest in multimodal large language model (MLLM) research. This line of research focuses on developing general-purpose LLMs through fine-tuning pre-trained LLMs and vision models. However, catastrophic forgetting, a notorious phenomenon where the fine-tuned model fails to retain similar performance compared to the pre-trained model, still remains an inherent problem in multimodal LLMs (MLLM). In this paper, we introduce EMT: Evaluating MulTimodality for evaluating the catastrophic forgetting in MLLMs, by treating each MLLM as an image classifier. We first apply EMT to evaluate several open-source fine-tuned MLLMs and we discover that almost all evaluated MLLMs fail to retain the same performance levels as their vision encoders on standard image classification tasks. Moreover, we continue fine-tuning LLaVA, an MLLM and utilize EMT to assess performance throughout the fine-tuning. Interestingly, our results suggest that early-stage fine-tuning on an image dataset improves performance across other image datasets, by enhancing the alignment of text and visual features. However, as fine-tuning proceeds, the MLLMs begin to hallucinate, resulting in a significant loss of generalizability, even when the image encoder remains frozen. Our results suggest that MLLMs have yet to demonstrate performance on par with their vision models on standard image classification tasks and the current MLLM fine-tuning procedure still has room for improvement. | http://arxiv.org/abs/2309.10313v4 | "2023-09-19T04:51:13Z" | cs.CL, cs.AI, cs.LG | 2,023 |
Leveraging Speech PTM, Text LLM, and Emotional TTS for Speech Emotion Recognition | Ziyang Ma, Wen Wu, Zhisheng Zheng, Yiwei Guo, Qian Chen, Shiliang Zhang, Xie Chen | In this paper, we explored how to boost speech emotion recognition (SER) with the state-of-the-art speech pre-trained model (PTM), data2vec, text generation technique, GPT-4, and speech synthesis technique, Azure TTS. First, we investigated the representation ability of different speech self-supervised pre-trained models, and we found that data2vec has a good representation ability on the SER task. Second, we employed a powerful large language model (LLM), GPT-4, and emotional text-to-speech (TTS) model, Azure TTS, to generate emotionally congruent text and speech. We carefully designed the text prompt and dataset construction, to obtain the synthetic emotional speech data with high quality. Third, we studied different ways of data augmentation to promote the SER task with synthetic speech, including random mixing, adversarial training, transfer learning, and curriculum learning. Experiments and ablation studies on the IEMOCAP dataset demonstrate the effectiveness of our method, compared with other data augmentation methods, and data augmentation with other synthetic data. | http://arxiv.org/abs/2309.10294v1 | "2023-09-19T03:52:01Z" | cs.CL, cs.AI, cs.SD, eess.AS | 2,023 |
Natural Language Dataset Generation Framework for Visualizations Powered by Large Language Models | Hyung-Kwon Ko, Hyeon Jeon, Gwanmo Park, Dae Hyun Kim, Nam Wook Kim, Juho Kim, Jinwook Seo | We introduce VL2NL, a Large Language Model (LLM) framework that generates rich and diverse NL datasets using only Vega-Lite specifications as input, thereby streamlining the development of Natural Language Interfaces (NLIs) for data visualization. To synthesize relevant chart semantics accurately and enhance syntactic diversity in each NL dataset, we leverage 1) a guided discovery incorporated into prompting so that LLMs can steer themselves to create faithful NL datasets in a self-directed manner; 2) a score-based paraphrasing to augment NL syntax along with four language axes. We also present a new collection of 1,981 real-world Vega-Lite specifications that have increased diversity and complexity than existing chart collections. When tested on our chart collection, VL2NL extracted chart semantics and generated L1/L2 captions with 89.4% and 76.0% accuracy, respectively. It also demonstrated generating and paraphrasing utterances and questions with greater diversity compared to the benchmarks. Last, we discuss how our NL datasets and framework can be utilized in real-world scenarios. The codes and chart collection are available at https://github.com/hyungkwonko/chart-llm. | http://arxiv.org/abs/2309.10245v4 | "2023-09-19T01:53:01Z" | cs.HC | 2,023 |
PolicyGPT: Automated Analysis of Privacy Policies with Large Language Models | Chenhao Tang, Zhengliang Liu, Chong Ma, Zihao Wu, Yiwei Li, Wei Liu, Dajiang Zhu, Quanzheng Li, Xiang Li, Tianming Liu, Lei Fan | Privacy policies serve as the primary conduit through which online service providers inform users about their data collection and usage procedures. However, in a bid to be comprehensive and mitigate legal risks, these policy documents are often quite verbose. In practical use, users tend to click the Agree button directly rather than reading them carefully. This practice exposes users to risks of privacy leakage and legal issues. Recently, the advent of Large Language Models (LLM) such as ChatGPT and GPT-4 has opened new possibilities for text analysis, especially for lengthy documents like privacy policies. In this study, we investigate a privacy policy text analysis framework PolicyGPT based on the LLM. This framework was tested using two datasets. The first dataset comprises of privacy policies from 115 websites, which were meticulously annotated by legal experts, categorizing each segment into one of 10 classes. The second dataset consists of privacy policies from 304 popular mobile applications, with each sentence manually annotated and classified into one of another 10 categories. Under zero-shot learning conditions, PolicyGPT demonstrated robust performance. For the first dataset, it achieved an accuracy rate of 97%, while for the second dataset, it attained an 87% accuracy rate, surpassing that of the baseline machine learning and neural network models. | http://arxiv.org/abs/2309.10238v1 | "2023-09-19T01:22:42Z" | cs.CL | 2,023 |
Stabilizing RLHF through Advantage Model and Selective Rehearsal | Baolin Peng, Linfeng Song, Ye Tian, Lifeng Jin, Haitao Mi, Dong Yu | Large Language Models (LLMs) have revolutionized natural language processing, yet aligning these models with human values and preferences using RLHF remains a significant challenge. This challenge is characterized by various instabilities, such as reward hacking and catastrophic forgetting. In this technical report, we propose two innovations to stabilize RLHF training: 1) Advantage Model, which directly models advantage score i.e., extra reward compared to the expected rewards and regulates score distributions across tasks to prevent reward hacking. 2) Selective Rehearsal, which mitigates catastrophic forgetting by strategically selecting data for PPO training and knowledge rehearsing. Our experimental analysis on public and proprietary datasets reveals that the proposed methods not only increase stability in RLHF training but also achieve higher reward scores and win rates. | http://arxiv.org/abs/2309.10202v1 | "2023-09-18T23:06:32Z" | cs.CL, cs.AI | 2,023 |
Positive and Risky Message Assessment for Music Products | Yigeng Zhang, Mahsa Shafaei, Fabio A. González, Thamar Solorio | In this work, we introduce a pioneering research challenge: evaluating positive and potentially harmful messages within music products. We initiate by setting a multi-faceted, multi-task benchmark for music content assessment. Subsequently, we introduce an efficient multi-task predictive model fortified with ordinality-enforcement to address this challenge. Our findings reveal that the proposed method not only significantly outperforms robust task-specific alternatives but also possesses the capability to assess multiple aspects simultaneously. Furthermore, through detailed case studies, where we employed Large Language Models (LLMs) as surrogates for content assessment, we provide valuable insights to inform and guide future research on this topic. The code for dataset creation and model implementation is publicly available at https://github.com/RiTUAL-UH/music-message-assessment. | http://arxiv.org/abs/2309.10182v2 | "2023-09-18T22:20:13Z" | cs.CL, cs.AI | 2,023 |
SMART-LLM: Smart Multi-Agent Robot Task Planning using Large Language Models | Shyam Sundar Kannan, Vishnunandan L. N. Venkatesh, Byung-Cheol Min | In this work, we introduce SMART-LLM, an innovative framework designed for embodied multi-robot task planning. SMART-LLM: Smart Multi-Agent Robot Task Planning using Large Language Models (LLMs), harnesses the power of LLMs to convert high-level task instructions provided as input into a multi-robot task plan. It accomplishes this by executing a series of stages, including task decomposition, coalition formation, and task allocation, all guided by programmatic LLM prompts within the few-shot prompting paradigm. We create a benchmark dataset designed for validating the multi-robot task planning problem, encompassing four distinct categories of high-level instructions that vary in task complexity. Our evaluation experiments span both simulation and real-world scenarios, demonstrating that the proposed model can achieve promising results for generating multi-robot task plans. The experimental videos, code, and datasets from the work can be found at https://sites.google.com/view/smart-llm/. | http://arxiv.org/abs/2309.10062v2 | "2023-09-18T18:17:56Z" | cs.RO | 2,023 |
Corpus Synthesis for Zero-shot ASR domain Adaptation using Large Language Models | Hsuan Su, Ting-Yao Hu, Hema Swetha Koppula, Raviteja Vemulapalli, Jen-Hao Rick Chang, Karren Yang, Gautam Varma Mantena, Oncel Tuzel | While Automatic Speech Recognition (ASR) systems are widely used in many real-world applications, they often do not generalize well to new domains and need to be finetuned on data from these domains. However, target-domain data usually are not readily available in many scenarios. In this paper, we propose a new strategy for adapting ASR models to new target domains without any text or speech from those domains. To accomplish this, we propose a novel data synthesis pipeline that uses a Large Language Model (LLM) to generate a target domain text corpus, and a state-of-the-art controllable speech synthesis model to generate the corresponding speech. We propose a simple yet effective in-context instruction finetuning strategy to increase the effectiveness of LLM in generating text corpora for new domains. Experiments on the SLURP dataset show that the proposed method achieves an average relative word error rate improvement of $28\%$ on unseen target domains without any performance drop in source domains. | http://arxiv.org/abs/2309.10707v1 | "2023-09-18T15:43:08Z" | eess.AS, cs.CL, cs.LG, cs.SD | 2,023 |
Efficient Avoidance of Vulnerabilities in Auto-completed Smart Contract Code Using Vulnerability-constrained Decoding | André Storhaug, Jingyue Li, Tianyuan Hu | Auto-completing code enables developers to speed up coding significantly. Recent advances in transformer-based large language model (LLM) technologies have been applied to code synthesis. However, studies show that many of such synthesized codes contain vulnerabilities. We propose a novel vulnerability-constrained decoding approach to reduce the amount of vulnerable code generated by such models. Using a small dataset of labeled vulnerable lines of code, we fine-tune an LLM to include vulnerability labels when generating code, acting as an embedded classifier. Then, during decoding, we deny the model to generate these labels to avoid generating vulnerable code. To evaluate the method, we chose to automatically complete Ethereum Blockchain smart contracts (SCs) as the case study due to the strict requirements of SC security. We first fine-tuned the 6-billion-parameter GPT-J model using 186,397 Ethereum SCs after removing the duplication from 2,217,692 SCs. The fine-tuning took more than one week using ten GPUs. The results showed that our fine-tuned model could synthesize SCs with an average BLEU (BiLingual Evaluation Understudy) score of 0.557. However, many codes in the auto-completed SCs were vulnerable. Using the code before the vulnerable line of 176 SCs containing different types of vulnerabilities to auto-complete the code, we found that more than 70% of the auto-completed codes were insecure. Thus, we further fine-tuned the model on other 941 vulnerable SCs containing the same types of vulnerabilities and applied vulnerability-constrained decoding. The fine-tuning took only one hour with four GPUs. We then auto-completed the 176 SCs again and found that our approach could identify 62% of the code to be generated as vulnerable and avoid generating 67% of them, indicating the approach could efficiently and effectively avoid vulnerabilities in the auto-completed code. | http://arxiv.org/abs/2309.09826v2 | "2023-09-18T14:47:34Z" | cs.CR, cs.AI, cs.CL | 2,023 |
LLM4Jobs: Unsupervised occupation extraction and standardization leveraging Large Language Models | Nan Li, Bo Kang, Tijl De Bie | Automated occupation extraction and standardization from free-text job postings and resumes are crucial for applications like job recommendation and labor market policy formation. This paper introduces LLM4Jobs, a novel unsupervised methodology that taps into the capabilities of large language models (LLMs) for occupation coding. LLM4Jobs uniquely harnesses both the natural language understanding and generation capacities of LLMs. Evaluated on rigorous experimentation on synthetic and real-world datasets, we demonstrate that LLM4Jobs consistently surpasses unsupervised state-of-the-art benchmarks, demonstrating its versatility across diverse datasets and granularities. As a side result of our work, we present both synthetic and real-world datasets, which may be instrumental for subsequent research in this domain. Overall, this investigation highlights the promise of contemporary LLMs for the intricate task of occupation extraction and standardization, laying the foundation for a robust and adaptable framework relevant to both research and industrial contexts. | http://arxiv.org/abs/2309.09708v2 | "2023-09-18T12:22:00Z" | cs.CL, cs.AI | 2,023 |
Fabricator: An Open Source Toolkit for Generating Labeled Training Data with Teacher LLMs | Jonas Golde, Patrick Haller, Felix Hamborg, Julian Risch, Alan Akbik | Most NLP tasks are modeled as supervised learning and thus require labeled training data to train effective models. However, manually producing such data at sufficient quality and quantity is known to be costly and time-intensive. Current research addresses this bottleneck by exploring a novel paradigm called zero-shot learning via dataset generation. Here, a powerful LLM is prompted with a task description to generate labeled data that can be used to train a downstream NLP model. For instance, an LLM might be prompted to "generate 500 movie reviews with positive overall sentiment, and another 500 with negative sentiment." The generated data could then be used to train a binary sentiment classifier, effectively leveraging an LLM as a teacher to a smaller student model. With this demo, we introduce Fabricator, an open-source Python toolkit for dataset generation. Fabricator implements common dataset generation workflows, supports a wide range of downstream NLP tasks (such as text classification, question answering, and entity recognition), and is integrated with well-known libraries to facilitate quick experimentation. With Fabricator, we aim to support researchers in conducting reproducible dataset generation experiments using LLMs and help practitioners apply this approach to train models for downstream tasks. | http://arxiv.org/abs/2309.09582v2 | "2023-09-18T08:45:47Z" | cs.CL, cs.AI | 2,023 |
Summarization is (Almost) Dead | Xiao Pu, Mingqi Gao, Xiaojun Wan | How well can large language models (LLMs) generate summaries? We develop new datasets and conduct human evaluation experiments to evaluate the zero-shot generation capability of LLMs across five distinct summarization tasks. Our findings indicate a clear preference among human evaluators for LLM-generated summaries over human-written summaries and summaries generated by fine-tuned models. Specifically, LLM-generated summaries exhibit better factual consistency and fewer instances of extrinsic hallucinations. Due to the satisfactory performance of LLMs in summarization tasks (even surpassing the benchmark of reference summaries), we believe that most conventional works in the field of text summarization are no longer necessary in the era of LLMs. However, we recognize that there are still some directions worth exploring, such as the creation of novel datasets with higher quality and more reliable evaluation methods. | http://arxiv.org/abs/2309.09558v1 | "2023-09-18T08:13:01Z" | cs.CL | 2,023 |
LayoutNUWA: Revealing the Hidden Layout Expertise of Large Language Models | Zecheng Tang, Chenfei Wu, Juntao Li, Nan Duan | Graphic layout generation, a growing research field, plays a significant role in user engagement and information perception. Existing methods primarily treat layout generation as a numerical optimization task, focusing on quantitative aspects while overlooking the semantic information of layout, such as the relationship between each layout element. In this paper, we propose LayoutNUWA, the first model that treats layout generation as a code generation task to enhance semantic information and harness the hidden layout expertise of large language models~(LLMs). More concretely, we develop a Code Instruct Tuning (CIT) approach comprising three interconnected modules: 1) the Code Initialization (CI) module quantifies the numerical conditions and initializes them as HTML code with strategically placed masks; 2) the Code Completion (CC) module employs the formatting knowledge of LLMs to fill in the masked portions within the HTML code; 3) the Code Rendering (CR) module transforms the completed code into the final layout output, ensuring a highly interpretable and transparent layout generation procedure that directly maps code to a visualized layout. We attain significant state-of-the-art performance (even over 50\% improvements) on multiple datasets, showcasing the strong capabilities of LayoutNUWA. Our code is available at https://github.com/ProjectNUWA/LayoutNUWA. | http://arxiv.org/abs/2309.09506v2 | "2023-09-18T06:35:10Z" | cs.CV, cs.CL | 2,023 |
CulturaX: A Cleaned, Enormous, and Multilingual Dataset for Large Language Models in 167 Languages | Thuat Nguyen, Chien Van Nguyen, Viet Dac Lai, Hieu Man, Nghia Trung Ngo, Franck Dernoncourt, Ryan A. Rossi, Thien Huu Nguyen | The driving factors behind the development of large language models (LLMs) with impressive learning capabilities are their colossal model sizes and extensive training datasets. Along with the progress in natural language processing, LLMs have been frequently made accessible to the public to foster deeper investigation and applications. However, when it comes to training datasets for these LLMs, especially the recent state-of-the-art models, they are often not fully disclosed. Creating training data for high-performing LLMs involves extensive cleaning and deduplication to ensure the necessary level of quality. The lack of transparency for training data has thus hampered research on attributing and addressing hallucination and bias issues in LLMs, hindering replication efforts and further advancements in the community. These challenges become even more pronounced in multilingual learning scenarios, where the available multilingual text datasets are often inadequately collected and cleaned. Consequently, there is a lack of open-source and readily usable dataset to effectively train LLMs in multiple languages. To overcome this issue, we present CulturaX, a substantial multilingual dataset with 6.3 trillion tokens in 167 languages, tailored for LLM development. Our dataset undergoes meticulous cleaning and deduplication through a rigorous pipeline of multiple stages to accomplish the best quality for model training, including language identification, URL-based filtering, metric-based cleaning, document refinement, and data deduplication. CulturaX is fully released to the public in HuggingFace to facilitate research and advancements in multilingual LLMs: https://huggingface.co/datasets/uonlp/CulturaX. | http://arxiv.org/abs/2309.09400v1 | "2023-09-17T23:49:10Z" | cs.CL, cs.AI | 2,023 |
Augmenting text for spoken language understanding with Large Language Models | Roshan Sharma, Suyoun Kim, Daniel Lazar, Trang Le, Akshat Shrivastava, Kwanghoon Ahn, Piyush Kansal, Leda Sari, Ozlem Kalinli, Michael Seltzer | Spoken semantic parsing (SSP) involves generating machine-comprehensible parses from input speech. Training robust models for existing application domains represented in training data or extending to new domains requires corresponding triplets of speech-transcript-semantic parse data, which is expensive to obtain. In this paper, we address this challenge by examining methods that can use transcript-semantic parse data (unpaired text) without corresponding speech. First, when unpaired text is drawn from existing textual corpora, Joint Audio Text (JAT) and Text-to-Speech (TTS) are compared as ways to generate speech representations for unpaired text. Experiments on the STOP dataset show that unpaired text from existing and new domains improves performance by 2% and 30% in absolute Exact Match (EM) respectively. Second, we consider the setting when unpaired text is not available in existing textual corpora. We propose to prompt Large Language Models (LLMs) to generate unpaired text for existing and new domains. Experiments show that examples and words that co-occur with intents can be used to generate unpaired text with Llama 2.0. Using the generated text with JAT and TTS for spoken semantic parsing improves EM on STOP by 1.4% and 2.6% absolute for existing and new domains respectively. | http://arxiv.org/abs/2309.09390v1 | "2023-09-17T22:25:34Z" | cs.CL, cs.SD, eess.AS | 2,023 |
Embrace Divergence for Richer Insights: A Multi-document Summarization Benchmark and a Case Study on Summarizing Diverse Information from News Articles | Kung-Hsiang Huang, Philippe Laban, Alexander R. Fabbri, Prafulla Kumar Choubey, Shafiq Joty, Caiming Xiong, Chien-Sheng Wu | Previous research in multi-document news summarization has typically concentrated on collating information that all sources agree upon. However, the summarization of diverse information dispersed across multiple articles about an event remains underexplored. In this paper, we propose a new task of summarizing diverse information encountered in multiple news articles encompassing the same event. To facilitate this task, we outlined a data collection schema for identifying diverse information and curated a dataset named DiverseSumm. The dataset includes 245 news stories, with each story comprising 10 news articles and paired with a human-validated reference. Next, to enable consistent automatic evaluation, we conducted a comprehensive analysis to pinpoint the position and verbosity biases when utilizing Large Language Model (LLM)-based metrics for evaluating the coverage and faithfulness of summaries. Through correlation analyses, we outline the best practices for effectively using automatic LLM-based metrics on the DiverseSumm dataset. Finally, we study how LLMs summarize multiple news articles by analyzing which type of diverse information LLMs are capable of identifying. Our analyses suggest that despite the extraordinary capabilities of LLMs in single-document summarization, the proposed task remains a complex challenge for them mainly due to their limited coverage, with GPT-4 only able to cover under 40% of the diverse information on average. | http://arxiv.org/abs/2309.09369v2 | "2023-09-17T20:28:17Z" | cs.CL | 2,023 |
Performance of the Pre-Trained Large Language Model GPT-4 on Automated Short Answer Grading | Gerd Kortemeyer | Automated Short Answer Grading (ASAG) has been an active area of machine-learning research for over a decade. It promises to let educators grade and give feedback on free-form responses in large-enrollment courses in spite of limited availability of human graders. Over the years, carefully trained models have achieved increasingly higher levels of performance. More recently, pre-trained Large Language Models (LLMs) emerged as a commodity, and an intriguing question is how a general-purpose tool without additional training compares to specialized models. We studied the performance of GPT-4 on the standard benchmark 2-way and 3-way datasets SciEntsBank and Beetle, where in addition to the standard task of grading the alignment of the student answer with a reference answer, we also investigated withholding the reference answer. We found that overall, the performance of the pre-trained general-purpose GPT-4 LLM is comparable to hand-engineered models, but worse than pre-trained LLMs that had specialized training. | http://arxiv.org/abs/2309.09338v1 | "2023-09-17T18:04:34Z" | cs.CL | 2,023 |
OWL: A Large Language Model for IT Operations | Hongcheng Guo, Jian Yang, Jiaheng Liu, Liqun Yang, Linzheng Chai, Jiaqi Bai, Junran Peng, Xiaorong Hu, Chao Chen, Dongfeng Zhang, Xu Shi, Tieqiao Zheng, Liangfan Zheng, Bo Zhang, Ke Xu, Zhoujun Li | With the rapid development of IT operations, it has become increasingly crucial to efficiently manage and analyze large volumes of data for practical applications. The techniques of Natural Language Processing (NLP) have shown remarkable capabilities for various tasks, including named entity recognition, machine translation and dialogue systems. Recently, Large Language Models (LLMs) have achieved significant improvements across various NLP downstream tasks. However, there is a lack of specialized LLMs for IT operations. In this paper, we introduce the OWL, a large language model trained on our collected OWL-Instruct dataset with a wide range of IT-related information, where the mixture-of-adapter strategy is proposed to improve the parameter-efficient tuning across different domains or tasks. Furthermore, we evaluate the performance of our OWL on the OWL-Bench established by us and open IT-related benchmarks. OWL demonstrates superior performance results on IT tasks, which outperforms existing models by significant margins. Moreover, we hope that the findings of our work will provide more insights to revolutionize the techniques of IT operations with specialized LLMs. | http://arxiv.org/abs/2309.09298v1 | "2023-09-17T15:19:29Z" | cs.CL | 2,023 |
Leveraging Large Language Models for Sequential Recommendation | Jesse Harte, Wouter Zorgdrager, Panos Louridas, Asterios Katsifodimos, Dietmar Jannach, Marios Fragkoulis | Sequential recommendation problems have received increasing attention in research during the past few years, leading to the inception of a large variety of algorithmic approaches. In this work, we explore how large language models (LLMs), which are nowadays introducing disruptive effects in many AI-based applications, can be used to build or improve sequential recommendation approaches. Specifically, we devise and evaluate three approaches to leverage the power of LLMs in different ways. Our results from experiments on two datasets show that initializing the state-of-the-art sequential recommendation model BERT4Rec with embeddings obtained from an LLM improves NDCG by 15-20% compared to the vanilla BERT4Rec model. Furthermore, we find that a simple approach that leverages LLM embeddings for producing recommendations, can provide competitive performance by highlighting semantically related items. We publicly share the code and data of our experiments to ensure reproducibility. | http://arxiv.org/abs/2309.09261v1 | "2023-09-17T12:53:53Z" | cs.IR | 2,023 |
Can Large Language Models Understand Real-World Complex Instructions? | Qianyu He, Jie Zeng, Wenhao Huang, Lina Chen, Jin Xiao, Qianxi He, Xunzhe Zhou, Lida Chen, Xintao Wang, Yuncheng Huang, Haoning Ye, Zihan Li, Shisong Chen, Yikai Zhang, Zhouhong Gu, Jiaqing Liang, Yanghua Xiao | Large language models (LLMs) can understand human instructions, showing their potential for pragmatic applications beyond traditional NLP tasks. However, they still struggle with complex instructions, which can be either complex task descriptions that require multiple tasks and constraints, or complex input that contains long context, noise, heterogeneous information and multi-turn format. Due to these features, LLMs often ignore semantic constraints from task descriptions, generate incorrect formats, violate length or sample count constraints, and be unfaithful to the input text. Existing benchmarks are insufficient to assess LLMs' ability to understand complex instructions, as they are close-ended and simple. To bridge this gap, we propose CELLO, a benchmark for evaluating LLMs' ability to follow complex instructions systematically. We design eight features for complex instructions and construct a comprehensive evaluation dataset from real-world scenarios. We also establish four criteria and develop corresponding metrics, as current ones are inadequate, biased or too strict and coarse-grained. We compare the performance of representative Chinese-oriented and English-oriented models in following complex instructions through extensive experiments. Resources of CELLO are publicly available at https://github.com/Abbey4799/CELLO. | http://arxiv.org/abs/2309.09150v2 | "2023-09-17T04:18:39Z" | cs.CL, cs.AI | 2,023 |
RMDM: A Multilabel Fakenews Dataset for Vietnamese Evidence Verification | Hai-Long Nguyen, Thi-Kieu-Trang Pham, Thai-Son Le, Tan-Minh Nguyen, Thi-Hai-Yen Vuong, Ha-Thanh Nguyen | In this study, we present a novel and challenging multilabel Vietnamese dataset (RMDM) designed to assess the performance of large language models (LLMs), in verifying electronic information related to legal contexts, focusing on fake news as potential input for electronic evidence. The RMDM dataset comprises four labels: real, mis, dis, and mal, representing real information, misinformation, disinformation, and mal-information, respectively. By including these diverse labels, RMDM captures the complexities of differing fake news categories and offers insights into the abilities of different language models to handle various types of information that could be part of electronic evidence. The dataset consists of a total of 1,556 samples, with 389 samples for each label. Preliminary tests on the dataset using GPT-based and BERT-based models reveal variations in the models' performance across different labels, indicating that the dataset effectively challenges the ability of various language models to verify the authenticity of such information. Our findings suggest that verifying electronic information related to legal contexts, including fake news, remains a difficult problem for language models, warranting further attention from the research community to advance toward more reliable AI models for potential legal applications. | http://arxiv.org/abs/2309.09071v1 | "2023-09-16T18:35:08Z" | cs.CL, cs.AI | 2,023 |
Sorted LLaMA: Unlocking the Potential of Intermediate Layers of Large Language Models for Dynamic Inference | Parsa Kavehzadeh, Mojtaba Valipour, Marzieh Tahaei, Ali Ghodsi, Boxing Chen, Mehdi Rezagholizadeh | Large language models (LLMs) have revolutionized natural language processing (NLP) by excelling at understanding and generating human-like text. However, their widespread deployment can be prohibitively expensive. SortedNet is a recent training technique for enabling dynamic inference by leveraging the modularity in networks and sorting sub-models based on computation/accuracy in a nested manner. We extend SortedNet to generative NLP tasks, making large language models dynamic without any Pre-Training and by only replacing Standard Fine-Tuning (SFT) with Sorted Fine-Tuning (SoFT). Our approach boosts model efficiency, eliminating the need for multiple models for various scenarios during inference. We show that this approach can unlock the power of intermediate layers of transformers in generating the target output. Our sub-models remain integral components of the original model, minimizing storage requirements and transition costs between different computational/latency budgets. The efficacy of our proposed method was demonstrated by applying it to tune LLaMA 2 13B on the Stanford Alpaca dataset for instruction following and TriviaQA for closed-book question answering. Our results show the superior performance of sub-models in comparison to Standard Fine-Tuning and SFT+ICT (Early-Exit), all achieved with efficient tuning and without additional memory usage during inference. | http://arxiv.org/abs/2309.08968v2 | "2023-09-16T11:58:34Z" | cs.CL, cs.LG | 2,023 |
ODSum: New Benchmarks for Open Domain Multi-Document Summarization | Yijie Zhou, Kejian Shi, Wencai Zhang, Yixin Liu, Yilun Zhao, Arman Cohan | Open-domain Multi-Document Summarization (ODMDS) is a critical tool for condensing vast arrays of documents into coherent, concise summaries. With a more inter-related document set, there does not necessarily exist a correct answer for the retrieval, making it hard to measure the retrieving performance. We propose a rule-based method to process query-based document summarization datasets into ODMDS datasets. Based on this method, we introduce a novel dataset, ODSum, a sophisticated case with its document index interdependent and often interrelated. We tackle ODMDS with the \textit{retrieve-then-summarize} method, and the performance of a list of retrievers and summarizers is investigated. Through extensive experiments, we identify variances in evaluation metrics and provide insights into their reliability. We also found that LLMs suffer great performance loss from retrieving errors. We further experimented methods to improve the performance as well as investigate their robustness against imperfect retrieval. We will release our data and code at https://github.com/yale-nlp/ODSum. | http://arxiv.org/abs/2309.08960v1 | "2023-09-16T11:27:34Z" | cs.CL | 2,023 |
Monolingual or Multilingual Instruction Tuning: Which Makes a Better Alpaca | Pinzhen Chen, Shaoxiong Ji, Nikolay Bogoychev, Andrey Kutuzov, Barry Haddow, Kenneth Heafield | Foundational large language models (LLMs) can be instruction-tuned to perform open-domain question answering, facilitating applications like chat assistants. While such efforts are often carried out in a single language, we empirically analyze cost-efficient strategies for multilingual scenarios. Our study employs the Alpaca dataset and machine translations of it to form multilingual data, which is then used to tune LLMs through either low-rank adaptation or full-parameter training. Under a controlled computation budget, comparisons show that multilingual tuning is on par or better than tuning a model for each language. Furthermore, multilingual tuning with downsampled data can be as powerful and more robust. Our findings serve as a guide for expanding language support through instruction tuning. | http://arxiv.org/abs/2309.08958v2 | "2023-09-16T11:22:46Z" | cs.CL, cs.AI | 2,023 |
Cross-Lingual Knowledge Editing in Large Language Models | Jiaan Wang, Yunlong Liang, Zengkui Sun, Yuxuan Cao, Jiarong Xu | Knowledge editing aims to change language models' performance on several special cases (i.e., editing scope) by infusing the corresponding expected knowledge into them. With the recent advancements in large language models (LLMs), knowledge editing has been shown as a promising technique to adapt LLMs to new knowledge without retraining from scratch. However, most of the previous studies neglect the multi-lingual nature of some main-stream LLMs (e.g., LLaMA, ChatGPT and GPT-4), and typically focus on monolingual scenarios, where LLMs are edited and evaluated in the same language. As a result, it is still unknown the effect of source language editing on a different target language. In this paper, we aim to figure out this cross-lingual effect in knowledge editing. Specifically, we first collect a large-scale cross-lingual synthetic dataset by translating ZsRE from English to Chinese. Then, we conduct English editing on various knowledge editing methods covering different paradigms, and evaluate their performance in Chinese, and vice versa. To give deeper analyses of the cross-lingual effect, the evaluation includes four aspects, i.e., reliability, generality, locality and portability. Furthermore, we analyze the inconsistent behaviors of the edited models and discuss their specific challenges. | http://arxiv.org/abs/2309.08952v1 | "2023-09-16T11:07:52Z" | cs.CL, cs.AI | 2,023 |
Enhancing Large Language Model Induced Task-Oriented Dialogue Systems Through Look-Forward Motivated Goals | Zhiyuan Hu, Yue Feng, Yang Deng, Zekun Li, See-Kiong Ng, Anh Tuan Luu, Bryan Hooi | Recently, the development of large language models (LLMs) has been significantly enhanced the question answering and dialogue generation, and makes them become increasingly popular in current practical scenarios. While unlike the general dialogue system which emphasizes the semantic performance, the task-oriented dialogue (ToD) systems aim to achieve the dialogue goal efficiently and successfully in multiple turns. Unfortunately, existing LLM-induced ToD systems lack the direct reward toward the final goal and do not take account of the dialogue proactivity that can strengthen the dialogue efficiency. To fill these gaps, we introduce the ProToD (Proactively Goal-Driven LLM-Induced ToD) approach, which anticipates the future dialogue actions and incorporates the goal-oriented reward signal to enhance ToD systems. Additionally, we present a novel evaluation method that assesses ToD systems based on goal-driven dialogue simulations. This method allows us to gauge user satisfaction, system efficiency and successful rate while overcoming the limitations of current Information and Success metrics. Empirical experiments conducted on the MultiWoZ 2.1 dataset demonstrate that our model can achieve superior performance using only 10% of the data compared to previous end-to-end fully supervised models. This improvement is accompanied by enhanced user satisfaction and efficiency. | http://arxiv.org/abs/2309.08949v1 | "2023-09-16T10:56:00Z" | cs.CL | 2,023 |
Multimodal Multi-Hop Question Answering Through a Conversation Between Tools and Efficiently Finetuned Large Language Models | Hossein Rajabzadeh, Suyuchen Wang, Hyock Ju Kwon, Bang Liu | We employ a tool-interacting divide-and-conquer strategy enabling large language models (LLMs) to answer complex multimodal multi-hop questions. In particular, we harness the power of large language models to divide a given multimodal multi-hop question into unimodal single-hop sub-questions to be answered by the appropriate tool from a predefined set of tools. After all corresponding tools provide the LLM with their answers, the LLM generates the next relevant unimodal single-hop question. To increase the reasoning ability of LLMs, we prompt chatGPT to generate a tool-interacting divide-and-conquer dataset. This dataset is then used to efficiently finetune the corresponding LLM. To assess the effectiveness of this approach, we conduct an evaluation on two recently introduced complex question-answering datasets. The experimental analysis demonstrate substantial improvements over existing state-of-the-art solutions, indicating the efficacy and generality of our strategy | http://arxiv.org/abs/2309.08922v1 | "2023-09-16T08:22:22Z" | cs.CL | 2,023 |
Investigating Subtler Biases in LLMs: Ageism, Beauty, Institutional, and Nationality Bias in Generative Models | Mahammed Kamruzzaman, Md. Minul Islam Shovon, Gene Louis Kim | LLMs are increasingly powerful and widely used to assist users in a variety of tasks. This use risks the introduction of LLM biases to consequential decisions such as job hiring, human performance evaluation, and criminal sentencing. Bias in NLP systems along the lines of gender and ethnicity has been widely studied, especially for specific stereotypes (e.g., Asians are good at math). In this paper, we investigate bias along less-studied but still consequential, dimensions, such as age and beauty, measuring subtler correlated decisions that LLMs make between social groups and unrelated positive and negative attributes. We ask whether LLMs hold wide-reaching biases of positive or negative sentiment for specific social groups similar to the ``what is beautiful is good'' bias found in people in experimental psychology. We introduce a template-generated dataset of sentence completion tasks that asks the model to select the most appropriate attribute to complete an evaluative statement about a person described as a member of a specific social group. We also reverse the completion task to select the social group based on an attribute. We report the correlations that we find for 4 cutting-edge LLMs. This dataset can be used as a benchmark to evaluate progress in more generalized biases and the templating technique can be used to expand the benchmark with minimal additional human annotation. | http://arxiv.org/abs/2309.08902v2 | "2023-09-16T07:07:04Z" | cs.CL | 2,023 |
X-PARADE: Cross-Lingual Textual Entailment and Information Divergence across Paragraphs | Juan Diego Rodriguez, Katrin Erk, Greg Durrett | Understanding when two pieces of text convey the same information is a goal touching many subproblems in NLP, including textual entailment and fact-checking. This problem becomes more complex when those two pieces of text are in different languages. Here, we introduce X-PARADE (Cross-lingual Paragraph-level Analysis of Divergences and Entailments), the first cross-lingual dataset of paragraph-level information divergences. Annotators label a paragraph in a target language at the span level and evaluate it with respect to a corresponding paragraph in a source language, indicating whether a given piece of information is the same, new, or new but can be inferred. This last notion establishes a link with cross-language NLI. Aligned paragraphs are sourced from Wikipedia pages in different languages, reflecting real information divergences observed in the wild. Armed with our dataset, we investigate a diverse set of approaches for this problem, including token alignment from machine translation, textual entailment methods that localize their decisions, and prompting LLMs. Our results show that these methods vary in their capability to handle inferable information, but they all fall short of human performance. | http://arxiv.org/abs/2309.08873v2 | "2023-09-16T04:34:55Z" | cs.CL | 2,023 |
PDFTriage: Question Answering over Long, Structured Documents | Jon Saad-Falcon, Joe Barrow, Alexa Siu, Ani Nenkova, David Seunghyun Yoon, Ryan A. Rossi, Franck Dernoncourt | Large Language Models (LLMs) have issues with document question answering (QA) in situations where the document is unable to fit in the small context length of an LLM. To overcome this issue, most existing works focus on retrieving the relevant context from the document, representing them as plain text. However, documents such as PDFs, web pages, and presentations are naturally structured with different pages, tables, sections, and so on. Representing such structured documents as plain text is incongruous with the user's mental model of these documents with rich structure. When a system has to query the document for context, this incongruity is brought to the fore, and seemingly trivial questions can trip up the QA system. To bridge this fundamental gap in handling structured documents, we propose an approach called PDFTriage that enables models to retrieve the context based on either structure or content. Our experiments demonstrate the effectiveness of the proposed PDFTriage-augmented models across several classes of questions where existing retrieval-augmented LLMs fail. To facilitate further research on this fundamental problem, we release our benchmark dataset consisting of 900+ human-generated questions over 80 structured documents from 10 different categories of question types for document QA. Our code and datasets will be released soon on Github. | http://arxiv.org/abs/2309.08872v2 | "2023-09-16T04:29:05Z" | cs.CL, cs.AI, cs.LG | 2,023 |
S3-DST: Structured Open-Domain Dialogue Segmentation and State Tracking in the Era of LLMs | Sarkar Snigdha Sarathi Das, Chirag Shah, Mengting Wan, Jennifer Neville, Longqi Yang, Reid Andersen, Georg Buscher, Tara Safavi | The traditional Dialogue State Tracking (DST) problem aims to track user preferences and intents in user-agent conversations. While sufficient for task-oriented dialogue systems supporting narrow domain applications, the advent of Large Language Model (LLM)-based chat systems has introduced many real-world intricacies in open-domain dialogues. These intricacies manifest in the form of increased complexity in contextual interactions, extended dialogue sessions encompassing a diverse array of topics, and more frequent contextual shifts. To handle these intricacies arising from evolving LLM-based chat systems, we propose joint dialogue segmentation and state tracking per segment in open-domain dialogue systems. Assuming a zero-shot setting appropriate to a true open-domain dialogue system, we propose S3-DST, a structured prompting technique that harnesses Pre-Analytical Recollection, a novel grounding mechanism we designed for improving long context tracking. To demonstrate the efficacy of our proposed approach in joint segmentation and state tracking, we evaluate S3-DST on a proprietary anonymized open-domain dialogue dataset, as well as publicly available DST and segmentation datasets. Across all datasets and settings, S3-DST consistently outperforms the state-of-the-art, demonstrating its potency and robustness the next generation of LLM-based chat systems. | http://arxiv.org/abs/2309.08827v1 | "2023-09-16T00:59:23Z" | cs.CL, cs.AI | 2,023 |
Self-training Strategies for Sentiment Analysis: An Empirical Study | Haochen Liu, Sai Krishna Rallabandi, Yijing Wu, Parag Pravin Dakle, Preethi Raghavan | Sentiment analysis is a crucial task in natural language processing that involves identifying and extracting subjective sentiment from text. Self-training has recently emerged as an economical and efficient technique for developing sentiment analysis models by leveraging a small amount of labeled data and a large amount of unlabeled data. However, given a set of training data, how to utilize them to conduct self-training makes a significant difference in the final performance of the model. We refer to this methodology as the self-training strategy. In this paper, we present an empirical study of various self-training strategies for sentiment analysis. First, we investigate the influence of the self-training strategy and hyper-parameters on the performance of traditional small language models (SLMs) in various few-shot settings. Second, we also explore the feasibility of leveraging large language models (LLMs) to help self-training. We propose and empirically compare several self-training strategies with the intervention of LLMs. Extensive experiments are conducted on three real-world sentiment analysis datasets. | http://arxiv.org/abs/2309.08777v2 | "2023-09-15T21:42:46Z" | cs.CL | 2,023 |
MusiLingo: Bridging Music and Text with Pre-trained Language Models for Music Captioning and Query Response | Zihao Deng, Yinghao Ma, Yudong Liu, Rongchen Guo, Ge Zhang, Wenhu Chen, Wenhao Huang, Emmanouil Benetos | Large Language Models (LLMs) have shown immense potential in multimodal applications, yet the convergence of textual and musical domains remains not well-explored. To address this gap, we present MusiLingo, a novel system for music caption generation and music-related query responses. MusiLingo employs a single projection layer to align music representations from the pre-trained frozen music audio model MERT with a frozen LLM, bridging the gap between music audio and textual contexts. We train it on an extensive music caption dataset and fine-tune it with instructional data. Due to the scarcity of high-quality music Q&A datasets, we created the MusicInstruct (MI) dataset from captions in the MusicCaps datasets, tailored for open-ended music inquiries. Empirical evaluations demonstrate its competitive performance in generating music captions and composing music-related Q&A pairs. Our introduced dataset enables notable advancements beyond previous ones. | http://arxiv.org/abs/2309.08730v3 | "2023-09-15T19:31:40Z" | eess.AS, cs.AI, cs.CL, cs.MM, cs.SD | 2,023 |
Fake News Detectors are Biased against Texts Generated by Large Language Models | Jinyan Su, Terry Yue Zhuo, Jonibek Mansurov, Di Wang, Preslav Nakov | The spread of fake news has emerged as a critical challenge, undermining trust and posing threats to society. In the era of Large Language Models (LLMs), the capability to generate believable fake content has intensified these concerns. In this study, we present a novel paradigm to evaluate fake news detectors in scenarios involving both human-written and LLM-generated misinformation. Intriguingly, our findings reveal a significant bias in many existing detectors: they are more prone to flagging LLM-generated content as fake news while often misclassifying human-written fake news as genuine. This unexpected bias appears to arise from distinct linguistic patterns inherent to LLM outputs. To address this, we introduce a mitigation strategy that leverages adversarial training with LLM-paraphrased genuine news. The resulting model yielded marked improvements in detection accuracy for both human and LLM-generated news. To further catalyze research in this domain, we release two comprehensive datasets, \texttt{GossipCop++} and \texttt{PolitiFact++}, thus amalgamating human-validated articles with LLM-generated fake and real news. | http://arxiv.org/abs/2309.08674v1 | "2023-09-15T18:04:40Z" | cs.CL, cs.AI | 2,023 |
Are Multilingual LLMs Culturally-Diverse Reasoners? An Investigation into Multicultural Proverbs and Sayings | Chen Cecilia Liu, Fajri Koto, Timothy Baldwin, Iryna Gurevych | Large language models (LLMs) are highly adept at question answering and reasoning tasks, but when reasoning in a situational context, human expectations vary depending on the relevant cultural common ground. As languages are associated with diverse cultures, LLMs should also be culturally-diverse reasoners. In this paper, we study the ability of a wide range of state-of-the-art multilingual LLMs (mLLMs) to reason with proverbs and sayings in a conversational context. Our experiments reveal that: (1) mLLMs "know" limited proverbs and memorizing proverbs does not mean understanding them within a conversational context; (2) mLLMs struggle to reason with figurative proverbs and sayings, and when asked to select the wrong answer (instead of asking it to select the correct answer); and (3) there is a "culture gap" in mLLMs when reasoning about proverbs and sayings translated from other languages. We construct and release our evaluation dataset MAPS (MulticultrAl Proverbs and Sayings) for proverb understanding with conversational context for six different languages. | http://arxiv.org/abs/2309.08591v2 | "2023-09-15T17:45:28Z" | cs.CL | 2,023 |
Casteist but Not Racist? Quantifying Disparities in Large Language Model Bias between India and the West | Khyati Khandelwal, Manuel Tonneau, Andrew M. Bean, Hannah Rose Kirk, Scott A. Hale | Large Language Models (LLMs), now used daily by millions of users, can encode societal biases, exposing their users to representational harms. A large body of scholarship on LLM bias exists but it predominantly adopts a Western-centric frame and attends comparatively less to bias levels and potential harms in the Global South. In this paper, we quantify stereotypical bias in popular LLMs according to an Indian-centric frame and compare bias levels between the Indian and Western contexts. To do this, we develop a novel dataset which we call Indian-BhED (Indian Bias Evaluation Dataset), containing stereotypical and anti-stereotypical examples for caste and religion contexts. We find that the majority of LLMs tested are strongly biased towards stereotypes in the Indian context, especially as compared to the Western context. We finally investigate Instruction Prompting as a simple intervention to mitigate such bias and find that it significantly reduces both stereotypical and anti-stereotypical biases in the majority of cases for GPT-3.5. The findings of this work highlight the need for including more diverse voices when evaluating LLMs. | http://arxiv.org/abs/2309.08573v1 | "2023-09-15T17:38:41Z" | cs.CL, cs.CY | 2,023 |
Connecting Large Language Models with Evolutionary Algorithms Yields Powerful Prompt Optimizers | Qingyan Guo, Rui Wang, Junliang Guo, Bei Li, Kaitao Song, Xu Tan, Guoqing Liu, Jiang Bian, Yujiu Yang | Large Language Models (LLMs) excel in various tasks, but they rely on carefully crafted prompts that often demand substantial human effort. To automate this process, in this paper, we propose a novel framework for discrete prompt optimization, called EvoPrompt, which borrows the idea of evolutionary algorithms (EAs) as they exhibit good performance and fast convergence. To enable EAs to work on discrete prompts, which are natural language expressions that need to be coherent and human-readable, we connect LLMs with EAs. This approach allows us to simultaneously leverage the powerful language processing capabilities of LLMs and the efficient optimization performance of EAs. Specifically, abstaining from any gradients or parameters, EvoPrompt starts from a population of prompts and iteratively generates new prompts with LLMs based on the evolutionary operators, improving the population based on the development set. We optimize prompts for both closed- and open-source LLMs including GPT-3.5 and Alpaca, on 31 datasets covering language understanding, generation tasks, as well as BIG-Bench Hard (BBH) tasks. EvoPrompt significantly outperforms human-engineered prompts and existing methods for automatic prompt generation (e.g., up to 25% on BBH). Furthermore, EvoPrompt demonstrates that connecting LLMs with EAs creates synergies, which could inspire further research on the combination of LLMs and conventional algorithms. | http://arxiv.org/abs/2309.08532v2 | "2023-09-15T16:50:09Z" | cs.CL, cs.AI | 2,023 |
Adversarial Attacks on Tables with Entity Swap | Aneta Koleva, Martin Ringsquandl, Volker Tresp | The capabilities of large language models (LLMs) have been successfully applied in the context of table representation learning. The recently proposed tabular language models have reported state-of-the-art results across various tasks for table interpretation. However, a closer look into the datasets commonly used for evaluation reveals an entity leakage from the train set into the test set. Motivated by this observation, we explore adversarial attacks that represent a more realistic inference setup. Adversarial attacks on text have been shown to greatly affect the performance of LLMs, but currently, there are no attacks targeting tabular language models. In this paper, we propose an evasive entity-swap attack for the column type annotation (CTA) task. Our CTA attack is the first black-box attack on tables, where we employ a similarity-based sampling strategy to generate adversarial examples. The experimental results show that the proposed attack generates up to a 70% drop in performance. | http://arxiv.org/abs/2309.08650v1 | "2023-09-15T15:03:33Z" | cs.CL, cs.AI, cs.CR | 2,023 |
MAPLE: Mobile App Prediction Leveraging Large Language Model Embeddings | Yonchanok Khaokaew, Hao Xue, Flora D. Salim | In recent years, predicting mobile app usage has become increasingly important for areas like app recommendation, user behaviour analysis, and mobile resource management. Existing models, however, struggle with the heterogeneous nature of contextual data and the user cold start problem. This study introduces a novel prediction model, Mobile App Prediction Leveraging Large Language Model Embeddings (MAPLE), which employs Large Language Models (LLMs) and installed app similarity to overcome these challenges. MAPLE utilises the power of LLMs to process contextual data and discern intricate relationships within it effectively. Additionally, we explore the use of installed app similarity to address the cold start problem, facilitating the modelling of user preferences and habits, even for new users with limited historical data. In essence, our research presents MAPLE as a novel, potent, and practical approach to app usage prediction, making significant strides in resolving issues faced by existing models. MAPLE stands out as a comprehensive and effective solution, setting a new benchmark for more precise and personalised app usage predictions. In tests on two real-world datasets, MAPLE surpasses contemporary models in both standard and cold start scenarios. These outcomes validate MAPLE's capacity for precise app usage predictions and its resilience against the cold start problem. This enhanced performance stems from the model's proficiency in capturing complex temporal patterns and leveraging contextual information. As a result, MAPLE can potentially improve personalised mobile app usage predictions and user experiences markedly. | http://arxiv.org/abs/2309.08648v3 | "2023-09-15T13:15:54Z" | cs.CL, cs.AI | 2,023 |
Unleashing Potential of Evidence in Knowledge-Intensive Dialogue Generation | Xianjie Wu, Jian Yang, Tongliang Li, Di Liang, Shiwei Zhang, Yiyang Du, Zhoujun Li | Incorporating external knowledge into dialogue generation (KIDG) is crucial for improving the correctness of response, where evidence fragments serve as knowledgeable snippets supporting the factual dialogue replies. However, introducing irrelevant content often adversely impacts reply quality and easily leads to hallucinated responses. Prior work on evidence retrieval and integration in dialogue systems falls short of fully leveraging existing evidence since the model fails to locate useful fragments accurately and overlooks hidden evidence labels within the KIDG dataset. To fully Unleash the potential of evidence, we propose a framework to effectively incorporate Evidence in knowledge-Intensive Dialogue Generation (u-EIDG). Specifically, we introduce an automatic evidence generation framework that harnesses the power of Large Language Models (LLMs) to mine reliable evidence veracity labels from unlabeled data. By utilizing these evidence labels, we train a reliable evidence indicator to effectively identify relevant evidence from retrieved passages. Furthermore, we propose an evidence-augmented generator with an evidence-focused attention mechanism, which allows the model to concentrate on evidenced segments. Experimental results on MultiDoc2Dial demonstrate the efficacy of evidential label augmentation and refined attention mechanisms in improving model performance. Further analysis confirms that the proposed method outperforms other baselines (+3~+5 points) regarding coherence and factual consistency. | http://arxiv.org/abs/2309.08380v1 | "2023-09-15T13:13:30Z" | cs.CL | 2,023 |
Using Large Language Model to Solve and Explain Physics Word Problems Approaching Human Level | Jingzhe Ding, Yan Cen, Xinyuan Wei | Our work demonstrates that large language model (LLM) pre-trained on texts can not only solve pure math word problems, but also physics word problems, whose solution requires calculation and inference based on prior physical knowledge. We collect and annotate the first physics word problem dataset-PhysQA, which contains over 1000 junior high school physics word problems (covering Kinematics, Mass&Density, Mechanics, Heat, Electricity). Then we use OpenAI' s GPT3.5 to generate the answer of these problems and found that GPT3.5 could automatically solve 49.3% of the problems through zero-shot learning and 73.2% through few-shot learning. This result demonstrates that by using similar problems and their answers as prompt, LLM could solve elementary physics word problems approaching human level performance. In addition to solving problems, GPT3.5 can also summarize the knowledge or topics covered by the problems, provide relevant explanations, and generate new physics word problems based on the input. Our work is the first research to focus on the automatic solving, explanation, and generation of physics word problems across various types and scenarios, and we achieve an acceptable and state-of-the-art accuracy. This underscores the potential of LLMs for further applications in secondary education. | http://arxiv.org/abs/2309.08182v2 | "2023-09-15T06:13:06Z" | cs.CL, cs.AI, I.2.7 | 2,023 |
FedJudge: Federated Legal Large Language Model | Linan Yue, Qi Liu, Yichao Du, Weibo Gao, Ye Liu, Fangzhou Yao | Large Language Models (LLMs) have gained prominence in the field of Legal Intelligence, offering potential applications in assisting legal professionals and laymen. However, the centralized training of these Legal LLMs raises data privacy concerns, as legal data is distributed among various institutions containing sensitive individual information. This paper addresses this challenge by exploring the integration of Legal LLMs with Federated Learning (FL) methodologies. By employing FL, Legal LLMs can be fine-tuned locally on devices or clients, and their parameters are aggregated and distributed on a central server, ensuring data privacy without directly sharing raw data. However, computation and communication overheads hinder the full fine-tuning of LLMs under the FL setting. Moreover, the distribution shift of legal data reduces the effectiveness of FL methods. To this end, in this paper, we propose the first Federated Legal Large Language Model (FedJudge) framework, which fine-tunes Legal LLMs efficiently and effectively. Specifically, FedJudge utilizes parameter-efficient fine-tuning methods to update only a few additional parameters during the FL training. Besides, we explore the continual learning methods to preserve the global model's important parameters when training local clients to mitigate the problem of data shifts. Extensive experimental results on three real-world datasets clearly validate the effectiveness of FedJudge. Code is released at https://github.com/yuelinan/FedJudge. | http://arxiv.org/abs/2309.08173v3 | "2023-09-15T05:45:44Z" | cs.CL | 2,023 |
InvestLM: A Large Language Model for Investment using Financial Domain Instruction Tuning | Yi Yang, Yixuan Tang, Kar Yan Tam | We present a new financial domain large language model, InvestLM, tuned on LLaMA-65B (Touvron et al., 2023), using a carefully curated instruction dataset related to financial investment. Inspired by less-is-more-for-alignment (Zhou et al., 2023), we manually curate a small yet diverse instruction dataset, covering a wide range of financial related topics, from Chartered Financial Analyst (CFA) exam questions to SEC filings to Stackexchange quantitative finance discussions. InvestLM shows strong capabilities in understanding financial text and provides helpful responses to investment related questions. Financial experts, including hedge fund managers and research analysts, rate InvestLM's response as comparable to those of state-of-the-art commercial models (GPT-3.5, GPT-4 and Claude-2). Zero-shot evaluation on a set of financial NLP benchmarks demonstrates strong generalizability. From a research perspective, this work suggests that a high-quality domain specific LLM can be tuned using a small set of carefully curated instructions on a well-trained foundation model, which is consistent with the Superficial Alignment Hypothesis (Zhou et al., 2023). From a practical perspective, this work develops a state-of-the-art financial domain LLM with superior capability in understanding financial texts and providing helpful investment advice, potentially enhancing the work efficiency of financial professionals. We release the model parameters to the research community. | http://arxiv.org/abs/2309.13064v1 | "2023-09-15T02:59:31Z" | q-fin.GN, cs.AI, cs.CL, cs.LG | 2,023 |
Subsets and Splits