Authors: Zekai Chen, Weeden Daniel, Po-yu Chen, Francois Buet-Golfouse
Abstract: The advent of personalized content generation by LLMs presents a novel challenge: how to efficiently adapt text to meet individual preferences without the unsustainable demand of creating a unique model for each user. This study introduces an innovative online method that employs neural bandit algorithms to dynamically optimize soft instruction embeddings based on user feedback, enhancing the personalization of open-ended text generation by white-box LLMs. Through rigorous experimentation on various tasks, we demonstrate significant performance improvements over baseline strategies. NeuralTS, in particular, leads to substantial enhancements in personalized news headline generation, achieving up to a 62.9% improvement in terms of best ROUGE scores and up to 2.76% increase in LLM-agent evaluation against the baseline.
Authors: Alphaeus Dmonte, Marcos Zampieri, Kevin Lybarger, Massimiliano Albanese
Abstract: Politics is one of the most prevalent topics discussed on social media platforms, particularly during major election cycles, where users engage in conversations about candidates and electoral processes. Malicious actors may use this opportunity to disseminate misinformation to undermine trust in the electoral process. The emergence of Large Language Models (LLMs) exacerbates this issue by enabling malicious actors to generate misinformation at an unprecedented scale. Artificial intelligence (AI)-generated content is often indistinguishable from authentic user content, raising concerns about the integrity of information on social networks. In this paper, we present a novel taxonomy for characterizing election-related claims. This taxonomy provides an instrument for analyzing election-related claims, with granular categories related to jurisdiction, equipment, processes, and the nature of claims. We introduce ElectAI, a novel benchmark dataset that consists of 9,900 tweets, each labeled as human- or AI-generated. For AI-generated tweets, the specific LLM variant that produced them is specified. We annotated a subset of 1,550 tweets using the proposed taxonomy to capture the characteristics of election-related claims. We explored the capabilities of LLMs in extracting the taxonomy attributes and trained various machine learning models using ElectAI to distinguish between human- and AI-generated posts and identify the specific LLM variant.
Authors: Darren Edge, Ha Trinh, Newman Cheng, Joshua Bradley, Alex Chao, Apurva Mody, Steven Truitt, Jonathan Larson
Abstract: The use of retrieval-augmented generation (RAG) to retrieve relevant information from an external knowledge source enables large language models (LLMs) to answer questions over private and/or previously unseen document collections. However, RAG fails on global questions directed at an entire text corpus, such as "What are the main themes in the dataset?", since this is inherently a query-focused summarization (QFS) task, rather than an explicit retrieval task. Prior QFS methods, meanwhile, fail to scale to the quantities of text indexed by typical RAG systems. To combine the strengths of these contrasting methods, we propose a Graph RAG approach to question answering over private text corpora that scales with both the generality of user questions and the quantity of source text to be indexed. Our approach uses an LLM to build a graph-based text index in two stages: first to derive an entity knowledge graph from the source documents, then to pregenerate community summaries for all groups of closely-related entities. Given a question, each community summary is used to generate a partial response, before all partial responses are again summarized in a final response to the user. For a class of global sensemaking questions over datasets in the 1 million token range, we show that Graph RAG leads to substantial improvements over a na\"ive RAG baseline for both the comprehensiveness and diversity of generated answers. An open-source, Python-based implementation of both global and local Graph RAG approaches is forthcoming at https://aka.ms/graphrag.
URLs: https://aka.ms/graphrag.
Authors: Cheng Kang, Daniel Novak, Katerina Urbanova, Yuqing Cheng, Yong Hu
Abstract: Large language models (LLMs) have demonstrated impressive generalization capabilities on specific tasks with human-written instruction data. However, the limited quantity, diversity, and professional expertise of such instruction data raise concerns about the performance of LLMs in psychotherapy tasks when provided with domain-specific instructions. To address this, we firstly propose Domain-Specific Assistant Instructions based on AlexanderStreet therapy, and secondly, we use an adaption fine-tuning method and retrieval augmented generation method to improve pre-trained LLMs. Through quantitative evaluation of linguistic quality using automatic and human evaluation, we observe that pre-trained LLMs on Psychotherapy Assistant Instructions outperform state-of-the-art LLMs response baselines. Our Assistant-Instruction approach offers a half-annotation method to align pre-trained LLMs with instructions and provide pre-trained LLMs with more psychotherapy knowledge.
Authors: Jiaqing Yuan, Lin Pan, Chung-Wei Hang, Jiang Guo, Jiarong Jiang, Bonan Min, Patrick Ng, Zhiguo Wang
Abstract: Large language models (LLMs) have shown remarkable performance on a variety of NLP tasks, and are being rapidly adopted in a wide range of use cases. It is therefore of vital importance to holistically evaluate the factuality of their generated outputs, as hallucinations remain a challenging issue. In this work, we focus on assessing LLMs' ability to recall factual knowledge learned from pretraining, and the factors that affect this ability. To that end, we construct FACT-BENCH, a representative benchmark covering 20 domains, 134 property types, 3 answer types, and different knowledge popularity levels. We benchmark 31 models from 10 model families and provide a holistic assessment of their strengths and weaknesses. We observe that instruction-tuning hurts knowledge recall, as pretraining-only models consistently outperform their instruction-tuned counterparts, and positive effects of model scaling, as larger models outperform smaller ones for all model families. However, the best performance from GPT-4 still represents a large gap with the upper-bound. We additionally study the role of in-context exemplars using counterfactual demonstrations, which lead to significant degradation of factual knowledge recall for large models. By further decoupling model known and unknown knowledge, we find the degradation is attributed to exemplars that contradict a model's known knowledge, as well as the number of such exemplars. Lastly, we fine-tune LLaMA-7B in different settings of known and unknown knowledge. In particular, fine-tuning on a model's known knowledge is beneficial, and consistently outperforms fine-tuning on unknown and mixed knowledge. We will make our benchmark publicly available.
Authors: Cuong Nhat Ha, Shima Asaadi, Sanjeev Kumar Karn, Oladimeji Farri, Tobias Heimann, Thomas Runkler
Abstract: Vision-language models, while effective in general domains and showing strong performance in diverse multi-modal applications like visual question-answering (VQA), struggle to maintain the same level of effectiveness in more specialized domains, e.g., medical. We propose a medical vision-language model that integrates large vision and language models adapted for the medical domain. This model goes through three stages of parameter-efficient training using three separate biomedical and radiology multi-modal visual and text datasets. The proposed model achieves state-of-the-art performance on the SLAKE 1.0 medical VQA (MedVQA) dataset with an overall accuracy of 87.5% and demonstrates strong performance on another MedVQA dataset, VQA-RAD, achieving an overall accuracy of 73.2%.
Authors: Mojdeh Rahmanian, Seyed Mostafa Fakhrahmad, Seyedeh Zahra Mousavi
Abstract: Objective: Clinical trials are essential for advancing pharmaceutical interventions, but they face a bottleneck in selecting eligible participants. Although leveraging electronic health records (EHR) for recruitment has gained popularity, the complex nature of unstructured medical texts presents challenges in efficiently identifying participants. Natural Language Processing (NLP) techniques have emerged as a solution with a recent focus on transformer models. In this study, we aimed to evaluate the performance of a prompt-based large language model for the cohort selection task from unstructured medical notes collected in the EHR. Methods: To process the medical records, we selected the most related sentences of the records to the eligibility criteria needed for the trial. The SNOMED CT concepts related to each eligibility criterion were collected. Medical records were also annotated with MedCAT based on the SNOMED CT ontology. Annotated sentences including concepts matched with the criteria-relevant terms were extracted. A prompt-based large language model (Generative Pre-trained Transformer (GPT) in this study) was then used with the extracted sentences as the training set. To assess its effectiveness, we evaluated the model's performance using the dataset from the 2018 n2c2 challenge, which aimed to classify medical records of 311 patients based on 13 eligibility criteria through NLP techniques. Results: Our proposed model showed the overall micro and macro F measures of 0.9061 and 0.8060 which were among the highest scores achieved by the experiments performed with this dataset. Conclusion: The application of a prompt-based large language model in this study to classify patients based on eligibility criteria received promising scores. Besides, we proposed a method of extractive summarization with the aid of SNOMED CT ontology that can be also applied to other medical texts.
Authors: Diogo A. P. Nunes, Joana Ferreira-Gomes, Fani Neto, David Martins de Matos
Abstract: Objectives: This study aims to systematically review the literature on the computational processing of the language of pain, whether generated by patients or physicians, identifying current trends and challenges. Methods: Following the PRISMA guidelines, a comprehensive literature search was conducted to select relevant studies on the computational processing of the language of pain and answer pre-defined research questions. Data extraction and synthesis were performed to categorize selected studies according to their primary purpose and outcome, patient and pain population, textual data, computational methodology, and outcome targets. Results: Physician-generated language of pain, specifically from clinical notes, was the most used data. Tasks included patient diagnosis and triaging, identification of pain mentions, treatment response prediction, biomedical entity extraction, correlation of linguistic features with clinical states, and lexico-semantic analysis of pain narratives. Only one study included previous linguistic knowledge on pain utterances in their experimental setup. Most studies targeted their outcomes for physicians, either directly as clinical tools or as indirect knowledge. The least targeted stage of clinical pain care was self-management, in which patients are most involved. The least studied dimensions of pain were affective and sociocultural. Only two studies measured how physician performance on clinical tasks improved with the inclusion of the proposed algorithm. Discussion: This study found that future research should focus on analyzing patient-generated language of pain, developing patient-centered resources for self-management and patient-empowerment, exploring affective and sociocultural aspects of pain, and measuring improvements in physician performance when aided by the proposed tools.
Authors: Zhuoqun Li, Hongyu Lin, Tianshu Wang, Boxi Cao, Yaojie Lu, Weixiang Zhou, Hao Wang, Zhenyu Zeng, Le Sun, Xianpei Han
Abstract: Linking a claim to grounded references is a critical ability to fulfill human demands for authentic and reliable information. Current studies are limited to specific tasks like information retrieval or semantic matching, where the claim-reference relationships are unique and fixed, while the referential knowledge linking (RKL) in real-world can be much more diverse and complex. In this paper, we propose universal referential knowledge linking (URL), which aims to resolve diversified referential knowledge linking tasks by one unified model. To this end, we propose a LLM-driven task-instructed representation compression, as well as a multi-view learning approach, in order to effectively adapt the instruction following and semantic understanding abilities of LLMs to referential knowledge linking. Furthermore, we also construct a new benchmark to evaluate ability of models on referential knowledge linking tasks across different scenarios. Experiments demonstrate that universal RKL is challenging for existing approaches, while the proposed framework can effectively resolve the task across various scenarios, and therefore outperforms previous approaches by a large margin.
Authors: John Bauer, Chloe Kiddon, Eric Yeh, Alex Shan, Christopher D. Manning
Abstract: Searching dependency graphs and manipulating them can be a time consuming and challenging task to get right. We document Semgrex, a system for searching dependency graphs, and introduce Ssurgeon, a system for manipulating the output of Semgrex. The compact language used by these systems allows for easy command line or API processing of dependencies. Additionally, integration with publicly released toolkits in Java and Python allows for searching text relations and attributes over natural text.
Authors: Hyeonseok Moon, Seungyoon Lee, Seongtae Hong, Seungjun Lee, Chanjun Park, Heuiseok Lim
Abstract: Translating major language resources to build minor language resources becomes a widely-used approach. Particularly in translating complex data points composed of multiple components, it is common to translate each component separately. However, we argue that this practice often overlooks the interrelation between components within the same data point. To address this limitation, we propose a novel MT pipeline that considers the intra-data relation in implementing MT for training data. In our MT pipeline, all the components in a data point are concatenated to form a single translation sequence and subsequently reconstructed to the data components after translation. We introduce a Catalyst Statement (CS) to enhance the intra-data relation, and Indicator Token (IT) to assist the decomposition of a translated sequence into its respective data components. Through our approach, we have achieved a considerable improvement in translation quality itself, along with its effectiveness as training data. Compared with the conventional approach that translates each data component separately, our method yields better training data that enhances the performance of the trained model by 2.690 points for the web page ranking (WPR) task, and 0.845 for the question generation (QG) task in the XGLUE benchmark.
Authors: Zijie Wang, Farzana Rashid, Eduardo Blanco
Abstract: People often answer yes-no questions without explicitly saying yes, no, or similar polar keywords. Figuring out the meaning of indirect answers is challenging, even for large language models. In this paper, we investigate this problem working with dialogues from multiple domains. We present new benchmarks in three diverse domains: movie scripts, tennis interviews, and airline customer service. We present an approach grounded on distant supervision and blended training to quickly adapt to a new dialogue domain. Experimental results show that our approach is never detrimental and yields F1 improvements as high as 11-34%.
Authors: Saranya Krishnamoorthy, Ayush Singh, Shabnam Tafreshi
Abstract: Electronic health records (EHR) even though a boon for healthcare practitioners, are growing convoluted and longer every day. Sifting around these lengthy EHRs is taxing and becomes a cumbersome part of physician-patient interaction. Several approaches have been proposed to help alleviate this prevalent issue either via summarization or sectioning, however, only a few approaches have truly been helpful in the past. With the rise of automated methods, machine learning (ML) has shown promise in solving the task of identifying relevant sections in EHR. However, most ML methods rely on labeled data which is difficult to get in healthcare. Large language models (LLMs) on the other hand, have performed impressive feats in natural language processing (NLP), that too in a zero-shot manner, i.e. without any labeled data. To that end, we propose using LLMs to identify relevant section headers. We find that GPT-4 can effectively solve the task on both zero and few-shot settings as well as segment dramatically better than state-of-the-art methods. Additionally, we also annotate a much harder real world dataset and find that GPT-4 struggles to perform well, alluding to further research and harder benchmarks.
Authors: Wenlong Zhao, Debanjan Mondal, Niket Tandon, Danica Dillion, Kurt Gray, Yuling Gu
Abstract: The awareness of multi-cultural human values is critical to the ability of language models (LMs) to generate safe and personalized responses. However, this awareness of LMs has been insufficiently studied, since the computer science community lacks access to the large-scale real-world data about multi-cultural values. In this paper, we present WorldValuesBench, a globally diverse, large-scale benchmark dataset for the multi-cultural value prediction task, which requires a model to generate a rating response to a value question based on demographic contexts. Our dataset is derived from an influential social science project, World Values Survey (WVS), that has collected answers to hundreds of value questions (e.g., social, economic, ethical) from 94,728 participants worldwide. We have constructed more than 20 million examples of the type "(demographic attributes, value question) $\rightarrow$ answer" from the WVS responses. We perform a case study using our dataset and show that the task is challenging for strong open and closed-source models. On merely $11.1\%$, $25.0\%$, $72.2\%$, and $75.0\%$ of the questions, Alpaca-7B, Vicuna-7B-v1.5, Mixtral-8x7B-Instruct-v0.1, and GPT-3.5 Turbo can respectively achieve $<0.2$ Wasserstein 1-distance from the human normalized answer distributions. WorldValuesBench opens up new research avenues in studying limitations and opportunities in multi-cultural value awareness of LMs.
Authors: Stephen Bothwell, Brian DuSell, David Chiang, Brian Krostenko
Abstract: Computational historical linguistics seeks to systematically understand processes of sound change, including during periods at which little to no formal recording of language is attested. At the same time, few computational resources exist which deeply explore phonological and morphological connections between proto-languages and their descendants. This is particularly true for the family of Italic languages. To assist historical linguists in the study of Italic sound change, we introduce the Proto-Italic to Latin (PILA) dataset, which consists of roughly 3,000 pairs of forms from Proto-Italic and Latin. We provide a detailed description of how our dataset was created and organized. Then, we exhibit PILA's value in two ways. First, we present baseline results for PILA on a pair of traditional computational historical linguistics tasks. Second, we demonstrate PILA's capability for enhancing other historical-linguistic datasets through a dataset compatibility study.
Authors: Sri Harsha Dumpala, Aman Jaiswal, Chandramouli Sastry, Evangelos Milios, Sageev Oore, Hassan Sajjad
Abstract: Despite their remarkable successes, state-of-the-art language models face challenges in grasping certain important semantic details. This paper introduces the VISLA (Variance and Invariance to Semantic and Lexical Alterations) benchmark, designed to evaluate the semantic and lexical understanding of language models. VISLA presents a 3-way semantic (in)equivalence task with a triplet of sentences associated with an image, to evaluate both vision-language models (VLMs) and unimodal language models (ULMs). An evaluation involving 34 VLMs and 20 ULMs reveals surprising difficulties in distinguishing between lexical and semantic variations. Spatial semantics encoded by language models also appear to be highly sensitive to lexical information. Notably, text encoders of VLMs demonstrate greater sensitivity to semantic and lexical variations than unimodal text encoders. Our contributions include the unification of image-to-text and text-to-text retrieval tasks, an off-the-shelf evaluation without fine-tuning, and assessing LMs' semantic (in)variance in the presence of lexical alterations. The results highlight strengths and weaknesses across diverse vision and unimodal language models, contributing to a deeper understanding of their capabilities. % VISLA enables a rigorous evaluation, shedding light on language models' capabilities in handling semantic and lexical nuances. Data and code will be made available at https://github.com/Sri-Harsha/visla_benchmark.
Authors: Kabir Ahuja, Vidhisha Balachandran, Madhur Panwar, Tianxing He, Noah A. Smith, Navin Goyal, Yulia Tsvetkov
Abstract: Transformers trained on natural language data have been shown to learn its hierarchical structure and generalize to sentences with unseen syntactic structures without explicitly encoding any structural bias. In this work, we investigate sources of inductive bias in transformer models and their training that could cause such generalization behavior to emerge. We extensively experiment with transformer models trained on multiple synthetic datasets and with different training objectives and show that while other objectives e.g. sequence-to-sequence modeling, prefix language modeling, often failed to lead to hierarchical generalization, models trained with the language modeling objective consistently learned to generalize hierarchically. We then conduct pruning experiments to study how transformers trained with the language modeling objective encode hierarchical structure. When pruned, we find joint existence of subnetworks within the model with different generalization behaviors (subnetworks corresponding to hierarchical structure and linear order). Finally, we take a Bayesian perspective to further uncover transformers' preference for hierarchical generalization: We establish a correlation between whether transformers generalize hierarchically on a dataset and whether the simplest explanation of that dataset is provided by a hierarchical grammar compared to regular grammars exhibiting linear generalization.
Authors: Yukai Zhou, Wenjie Wang
Abstract: Ensuring the safety alignment of Large Language Models (LLMs) is crucial to generating responses consistent with human values. Despite their ability to recognize and avoid harmful queries, LLMs are vulnerable to "jailbreaking" attacks, where carefully crafted prompts elicit them to produce toxic content. One category of jailbreak attacks is reformulating the task as adversarial attacks by eliciting the LLM to generate an affirmative response. However, the typical attack in this category GCG has very limited attack success rate. In this study, to better study the jailbreak attack, we introduce the DSN (Don't Say No) attack, which prompts LLMs to not only generate affirmative responses but also novelly enhance the objective to suppress refusals. In addition, another challenge lies in jailbreak attacks is the evaluation, as it is difficult to directly and accurately assess the harmfulness of the attack. The existing evaluation such as refusal keyword matching has its own limitation as it reveals numerous false positive and false negative instances. To overcome this challenge, we propose an ensemble evaluation pipeline incorporating Natural Language Inference (NLI) contradiction assessment and two external LLM evaluators. Extensive experiments demonstrate the potency of the DSN and the effectiveness of ensemble evaluation compared to baseline methods.
Authors: Florian Pl\"otzky, Niklas Kiehne, Wolf-Tilo Balke
Abstract: Our world is shaped by events of various complexity. This includes both small-scale local events like local farmer markets and large complex events like political and military conflicts. The latter are typically not observed directly but through the lenses of intermediaries like newspapers or social media. In other words, we do not witness the unfolding of such events directly but are confronted with narratives surrounding them. Such narratives capture different aspects of a complex event and may also differ with respect to the narrator. Thus, they provide a rich semantics concerning real-world events. In this paper, we show how narratives concerning complex events can be constructed and utilized. We provide a formal representation of narratives based on recursive nodes to represent multiple levels of detail and discuss how narratives can be bound to event-centric knowledge graphs. Additionally, we provide an algorithm based on incremental prompting techniques that mines such narratives from texts to account for different perspectives on complex events. Finally, we show the effectiveness and future research directions in a proof of concept.
Authors: Xingchen Song, Di Wu, Binbin Zhang, Dinghao Zhou, Zhendong Peng, Bo Dang, Fuping Pan, Chao Yang
Abstract: Scale has opened new frontiers in natural language processing, but at a high cost. In response, by learning to only activate a subset of parameters in training and inference, Mixture-of-Experts (MoE) have been proposed as an energy efficient path to even larger and more capable language models and this shift towards a new generation of foundation models is gaining momentum, particularly within the field of Automatic Speech Recognition (ASR). Recent works that incorporating MoE into ASR models have complex designs such as routing frames via supplementary embedding network, improving multilingual ability for the experts, and utilizing dedicated auxiliary losses for either expert load balancing or specific language handling. We found that delicate designs are not necessary, while an embarrassingly simple substitution of MoE layers for all Feed-Forward Network (FFN) layers is competent for the ASR task. To be more specific, we benchmark our proposed model on a large scale inner-source dataset (160k hours), the results show that we can scale our baseline Conformer (Dense-225M) to its MoE counterparts (MoE-1B) and achieve Dense-1B level Word Error Rate (WER) while maintaining a Dense-225M level Real Time Factor (RTF). Furthermore, by applying Unified 2-pass framework with bidirectional attention decoders (U2++), we achieve the streaming and non-streaming decoding modes in a single MoE based model, which we call U2++ MoE. We hope that our study can facilitate the research on scaling speech foundation models without sacrificing deployment efficiency.
Authors: Md Nayem Uddin, Enfa Rose George, Eduardo Blanco, Steven Corman
Abstract: This paper presents a question-answering approach to extract document-level event-argument structures. We automatically ask and answer questions for each argument type an event may have. Questions are generated using manually defined templates and generative transformers. Template-based questions are generated using predefined role-specific wh-words and event triggers from the context document. Transformer-based questions are generated using large language models trained to formulate questions based on a passage and the expected answer. Additionally, we develop novel data augmentation strategies specialized in inter-sentential event-argument relations. We use a simple span-swapping technique, coreference resolution, and large language models to augment the training instances. Our approach enables transfer learning without any corpora-specific modifications and yields competitive results with the RAMS dataset. It outperforms previous work, and it is especially beneficial to extract arguments that appear in different sentences than the event trigger. We also present detailed quantitative and qualitative analyses shedding light on the most common errors made by our best model.
Authors: Changho Lee, Janghoon Han, Seonghyeon Ye, Stanley Jungkyu Choi, Honglak Lee, Kyunghoon Bae
Abstract: Instruction tuning has shown its ability to not only enhance zero-shot generalization across various tasks but also its effectiveness in improving the performance of specific tasks. A crucial aspect in instruction tuning for a particular task is a strategic selection of related tasks that offer meaningful supervision, thereby enhancing efficiency and preventing performance degradation from irrelevant tasks. Our research reveals that leveraging instruction information \textit{alone} enables the identification of pertinent tasks for instruction tuning. This approach is notably simpler compared to traditional methods that necessitate complex measurements of pairwise transferability between tasks or the creation of data samples for the target task. Furthermore, by additionally learning the unique instructional template style of the meta-dataset, we observe an improvement in task selection accuracy, which contributes to enhanced overall performance. Experimental results demonstrate that training on a small set of tasks, chosen solely based on the instructions, leads to substantial performance improvements on benchmarks like P3, Big-Bench, NIV2, and Big-Bench Hard. Significantly, these improvements exceed those achieved by prior task selection methods, highlighting the efficacy of our approach.
Authors: Zineddine Bettouche, Anas Safi, Andreas Fischer
Abstract: Managing the semantic quality of the categorization in large textual datasets, such as Wikipedia, presents significant challenges in terms of complexity and cost. In this paper, we propose leveraging transformer models to distill semantic information from texts in the Wikipedia dataset and its associated categories into a latent space. We then explore different approaches based on these encodings to assess and enhance the semantic identity of the categories. Our graphical approach is powered by Convex Hull, while we utilize Hierarchical Navigable Small Worlds (HNSWs) for the hierarchical approach. As a solution to the information loss caused by the dimensionality reduction, we modulate the following mathematical solution: an exponential decay function driven by the Euclidean distances between the high-dimensional encodings of the textual categories. This function represents a filter built around a contextual category and retrieves items with a certain Reconsideration Probability (RP). Retrieving high-RP items serves as a tool for database administrators to improve data groupings by providing recommendations and identifying outliers within a contextual framework.
Authors: Isablle Lorge, Dam W. Joyce, Andrey Kormilitzin
Abstract: Mental health in children and adolescents has been steadily deteriorating over the past few years [ 1 ]. The recent advent of Large Language Models (LLMs) offers much hope for cost and time efficient scaling of monitoring and intervention, yet despite specifically prevalent issues such as school bullying and eating disorders, previous studies on have not investigated performance in this domain or for open information extraction where the set of answers is not predetermined. We create a new dataset of Reddit posts from adolescents aged 12-19 annotated by expert psychiatrists for the following categories: TRAUMA, PRECARITY, CONDITION, SYMPTOMS, SUICIDALITY and TREATMENT and compare expert labels to annotations from two top performing LLMs (GPT3.5 and GPT4). In addition, we create two synthetic datasets to assess whether LLMs perform better when annotating data as they generate it. We find GPT4 to be on par with human inter-annotator agreement and performance on synthetic data to be substantially higher, however we find the model still occasionally errs on issues of negation and factuality and higher performance on synthetic data is driven by greater complexity of real data rather than inherent advantage.
Authors: Yash Saxena, Sarthak Chopra, Arunendra Mani Tripathi
Abstract: Large Language Models (LLMs) are extensively used today across various sectors, including academia, research, business, and finance, for tasks such as text generation, summarization, and translation. Despite their widespread adoption, these models often produce incorrect and misleading information, exhibiting a tendency to hallucinate. This behavior can be attributed to several factors, with consistency and reasoning capabilities being significant contributors. LLMs frequently lack the ability to generate explanations and engage in coherent reasoning, leading to inaccurate responses. Moreover, they exhibit inconsistencies in their outputs. This paper aims to evaluate and compare the consistency and reasoning capabilities of both public and proprietary LLMs. The experiments utilize the Boolq dataset as the ground truth, comprising questions, answers, and corresponding explanations. Queries from the dataset are presented as prompts to the LLMs, and the generated responses are evaluated against the ground truth answers. Additionally, explanations are generated to assess the models' reasoning abilities. Consistency is evaluated by repeatedly presenting the same query to the models and observing for variations in their responses. For measuring reasoning capabilities, the generated explanations are compared to the ground truth explanations using metrics such as BERT, BLEU, and F-1 scores. The findings reveal that proprietary models generally outperform public models in terms of both consistency and reasoning capabilities. However, even when presented with basic general knowledge questions, none of the models achieved a score of 90\% in both consistency and reasoning. This study underscores the direct correlation between consistency and reasoning abilities in LLMs and highlights the inherent reasoning challenges present in current language models.
Authors: Youmi Ma, An Wang, Naoaki Okazaki
Abstract: Document-level Relation Extraction (DocRE) is the task of extracting all semantic relationships from a document. While studies have been conducted on English DocRE, limited attention has been given to DocRE in non-English languages. This work delves into effectively utilizing existing English resources to promote DocRE studies in non-English languages, with Japanese as the representative case. As an initial attempt, we construct a dataset by transferring an English dataset to Japanese. However, models trained on such a dataset suffer from low recalls. We investigate the error cases and attribute the failure to different surface structures and semantics of documents translated from English and those written by native speakers. We thus switch to explore if the transferred dataset can assist human annotation on Japanese documents. In our proposal, annotators edit relation predictions from a model trained on the transferred dataset. Quantitative analysis shows that relation recommendations suggested by the model help reduce approximately 50% of the human edit steps compared with the previous approach. Experiments quantify the performance of existing DocRE models on our collected dataset, portraying the challenges of Japanese and cross-lingual DocRE.
Authors: Elizabeth Fons, Rachneet Kaur, Soham Palande, Zhen Zeng, Svitlana Vyetrenko, Tucker Balch
Abstract: Large Language Models (LLMs) offer the potential for automatic time series analysis and reporting, which is a critical task across many domains, spanning healthcare, finance, climate, energy, and many more. In this paper, we propose a framework for rigorously evaluating the capabilities of LLMs on time series understanding, encompassing both univariate and multivariate forms. We introduce a comprehensive taxonomy of time series features, a critical framework that delineates various characteristics inherent in time series data. Leveraging this taxonomy, we have systematically designed and synthesized a diverse dataset of time series, embodying the different outlined features. This dataset acts as a solid foundation for assessing the proficiency of LLMs in comprehending time series. Our experiments shed light on the strengths and limitations of state-of-the-art LLMs in time series understanding, revealing which features these models readily comprehend effectively and where they falter. In addition, we uncover the sensitivity of LLMs to factors including the formatting of the data, the position of points queried within a series and the overall time series length.
Authors: Ulme Wennberg, Gustav Eje Henter
Abstract: It has been found that Transformer-based language models have the ability to perform basic quantitative reasoning. In this paper, we propose a method for studying how these models internally represent numerical data, and use our proposal to analyze the ALBERT family of language models. Specifically, we extract the learned embeddings these models use to represent tokens that correspond to numbers and ordinals, and subject these embeddings to Principal Component Analysis (PCA). PCA results reveal that ALBERT models of different sizes, trained and initialized separately, consistently learn to use the axes of greatest variation to represent the approximate ordering of various numerical concepts. Numerals and their textual counterparts are represented in separate clusters, but increase along the same direction in 2D space. Our findings illustrate that language models, trained purely to model text, can intuit basic mathematical concepts, opening avenues for NLP applications that intersect with quantitative reasoning.
Authors: Zhihao Zhu, Ninglu Shao, Defu Lian, Chenwang Wu, Zheng Liu, Yi Yang, Enhong Chen
Abstract: Large language models (LLMs) show early signs of artificial general intelligence but struggle with hallucinations. One promising solution to mitigate these hallucinations is to store external knowledge as embeddings, aiding LLMs in retrieval-augmented generation. However, such a solution risks compromising privacy, as recent studies experimentally showed that the original text can be partially reconstructed from text embeddings by pre-trained language models. The significant advantage of LLMs over traditional pre-trained models may exacerbate these concerns. To this end, we investigate the effectiveness of reconstructing original knowledge and predicting entity attributes from these embeddings when LLMs are employed. Empirical findings indicate that LLMs significantly improve the accuracy of two evaluated tasks over those from pre-trained models, regardless of whether the texts are in-distribution or out-of-distribution. This underscores a heightened potential for LLMs to jeopardize user privacy, highlighting the negative consequences of their widespread use. We further discuss preliminary strategies to mitigate this risk.
Authors: Jianyu Zheng, Fengfei Fan, Jianquan Li
Abstract: Unsupervised cross-lingual transfer involves transferring knowledge between languages without explicit supervision. Although numerous studies have been conducted to improve performance in such tasks by focusing on cross-lingual knowledge, particularly lexical and syntactic knowledge, current approaches are limited as they only incorporate syntactic or lexical information. Since each type of information offers unique advantages and no previous attempts have combined both, we attempt to explore the potential of this approach. In this paper, we present a novel framework called "Lexicon-Syntax Enhanced Multilingual BERT" that combines both lexical and syntactic knowledge. Specifically, we use Multilingual BERT (mBERT) as the base model and employ two techniques to enhance its learning capabilities. The code-switching technique is used to implicitly teach the model lexical alignment information, while a syntactic-based graph attention network is designed to help the model encode syntactic structure. To integrate both types of knowledge, we input code-switched sequences into both the syntactic module and the mBERT base model simultaneously. Our extensive experimental results demonstrate this framework can consistently outperform all baselines of zero-shot cross-lingual transfer, with the gains of 1.0~3.7 points on text classification, named entity recognition (ner), and semantic parsing tasks. Keywords:cross-lingual transfer, lexicon, syntax, code-switching, graph attention network
Authors: Xiang Li, Yiqun Yao, Xin Jiang, Xuezhi Fang, Chao Wang, Xinzhang Liu, Zihan Wang, Yu Zhao, Xin Wang, Yuyao Huang, Shuangyong Song, Yongxiang Li, Zheng Zhang, Bo Zhao, Aixin Sun, Yequan Wang, Zhongjiang He, Zhongyuan Wang, Xuelong Li, Tiejun Huang
Abstract: Large language models (LLMs) have showcased profound capabilities in language understanding and generation, facilitating a wide array of applications. However, there is a notable paucity of detailed, open-sourced methodologies on efficiently scaling LLMs beyond 50 billion parameters with minimum trial-and-error cost and computational resources. In this report, we introduce Tele-FLM (aka FLM-2), a 52B open-sourced multilingual large language model that features a stable, efficient pre-training paradigm and enhanced factual judgment capabilities. Tele-FLM demonstrates superior multilingual language modeling abilities, measured by BPB on textual corpus. Besides, in both English and Chinese foundation model evaluation, it is comparable to strong open-sourced models that involve larger pre-training FLOPs, such as Llama2-70B and DeepSeek-67B. In addition to the model weights, we share the core designs, engineering practices, and training details, which we expect to benefit both the academic and industrial communities.
Authors: Lav\'inia de Carvalho Moraes, Irene Cristina Silv\'erio, Rafael Alexandre Sousa Marques, Bianca de Castro Anaia, Dandara Freitas de Paula, Maria Carolina Schincariol de Faria, Iury Cleveston, Alana de Santana Correia, Raquel Meister Ko Freitag
Abstract: Linguistic ambiguity continues to represent a significant challenge for natural language processing (NLP) systems, notwithstanding the advancements in architectures such as Transformers and BERT. Inspired by the recent success of instructional models like ChatGPT and Gemini (In 2023, the artificial intelligence was called Bard.), this study aims to analyze and discuss linguistic ambiguity within these models, focusing on three types prevalent in Brazilian Portuguese: semantic, syntactic, and lexical ambiguity. We create a corpus comprising 120 sentences, both ambiguous and unambiguous, for classification, explanation, and disambiguation. The models capability to generate ambiguous sentences was also explored by soliciting sets of sentences for each type of ambiguity. The results underwent qualitative analysis, drawing on recognized linguistic references, and quantitative assessment based on the accuracy of the responses obtained. It was evidenced that even the most sophisticated models, such as ChatGPT and Gemini, exhibit errors and deficiencies in their responses, with explanations often providing inconsistent. Furthermore, the accuracy peaked at 49.58 percent, indicating the need for descriptive studies for supervised learning.
Authors: Sangryul Kim, Donghee Han, Sehyun Kim
Abstract: Recently, deep learning-based language models have significantly enhanced text-to-SQL tasks, with promising applications in retrieving patient records within the medical domain. One notable challenge in such applications is discerning unanswerable queries. Through fine-tuning model, we demonstrate the feasibility of converting medical record inquiries into SQL queries. Additionally, we introduce an entropy-based method to identify and filter out unanswerable results. We further enhance result quality by filtering low-confidence SQL through log probability-based distribution, while grammatical and schema errors are mitigated by executing queries on the actual database. We experimentally verified that our method can filter unanswerable questions, which can be widely utilized even when the parameters of the model are not accessible, and that it can be effectively utilized in practice.
Authors: Lydia Uhler, Verena Jordan, J\"urgen Buder, Markus Huff, Frank Papenmeier
Abstract: We explored the addition bias, a cognitive tendency to prefer adding elements over removing them to alter an initial state or structure, by conducting four preregistered experiments examining the problem-solving behavior of both humans and OpenAl's GPT-4 large language model. The experiments involved 588 participants from the U.S. and 680 iterations of the GPT-4 model. The problem-solving task was either to create symmetry within a grid (Experiments 1 and 3) or to edit a summary (Experiments 2 and 4). As hypothesized, we found that overall, the addition bias was present. Solution efficiency (Experiments 1 and 2) and valence of the instruction (Experiments 3 and 4) played important roles. Human participants were less likely to use additive strategies when subtraction was relatively more efficient than when addition and subtraction were equally efficient. GPT-4 exhibited the opposite behavior, with a strong addition bias when subtraction was more efficient. In terms of instruction valence, GPT-4 was more likely to add words when asked to "improve" compared to "edit", whereas humans did not show this effect. When we looked at the addition bias under different conditions, we found more biased responses for GPT-4 compared to humans. Our findings highlight the importance of considering comparable and sometimes superior subtractive alternatives, as well as reevaluating one's own and particularly the language models' problem-solving behavior.
Authors: Giorgio Piatti, Zhijing Jin, Max Kleiman-Weiner, Bernhard Sch\"olkopf, Mrinmaya Sachan, Rada Mihalcea
Abstract: In the rapidly evolving field of artificial intelligence, ensuring safe decision-making of Large Language Models (LLMs) is a significant challenge. This paper introduces Governance of the Commons Simulation (GovSim), a simulation platform designed to study strategic interactions and cooperative decision-making in LLMs. Through this simulation environment, we explore the dynamics of resource sharing among AI agents, highlighting the importance of ethical considerations, strategic planning, and negotiation skills. GovSim is versatile and supports any text-based agent, including LLMs agents. Using the Generative Agent framework, we create a standard agent that facilitates the integration of different LLMs. Our findings reveal that within GovSim, only two out of 15 tested LLMs managed to achieve a sustainable outcome, indicating a significant gap in the ability of models to manage shared resources. Furthermore, we find that by removing the ability of agents to communicate, they overuse the shared resource, highlighting the importance of communication for cooperation. Interestingly, most LLMs lack the ability to make universalized hypotheses, which highlights a significant weakness in their reasoning skills. We open source the full suite of our research results, including the simulation environment, agent prompts, and a comprehensive web interface.
Authors: Mostafa Elhoushi, Akshat Shrivastava, Diana Liskovich, Basil Hosmer, Bram Wasti, Liangzhen Lai, Anas Mahmoud, Bilge Acun, Saurabh Agarwal, Ahmed Roman, Ahmed A Aly, Beidi Chen, Carole-Jean Wu
Abstract: We present LayerSkip, an end-to-end solution to speed-up inference of large language models (LLMs). First, during training we apply layer dropout, with low dropout rates for earlier layers and higher dropout rates for later layers, and an early exit loss where all transformer layers share the same exit. Second, during inference, we show that this training recipe increases the accuracy of early exit at earlier layers, without adding any auxiliary layers or modules to the model. Third, we present a novel self-speculative decoding solution where we exit at early layers and verify and correct with remaining layers of the model. Our proposed self-speculative decoding approach has less memory footprint than other speculative decoding approaches and benefits from shared compute and activations of the draft and verification stages. We run experiments on different Llama model sizes on different types of training: pretraining from scratch, continual pretraining, finetuning on specific data domain, and finetuning on specific task. We implement our inference solution and show speedups of up to 2.16x on summarization for CNN/DM documents, 1.82x on coding, and 2.0x on TOPv2 semantic parsing task.
Authors: Chanho Park, Mingjie Chen, Thomas Hain
Abstract: Word error rate (WER) is a metric used to evaluate the quality of transcriptions produced by Automatic Speech Recognition (ASR) systems. In many applications, it is of interest to estimate WER given a pair of a speech utterance and a transcript. Previous work on WER estimation focused on building models that are trained with a specific ASR system in mind (referred to as ASR system-dependent). These are also domain-dependent and inflexible in real-world applications. In this paper, a hypothesis generation method for ASR System-Independent WER estimation (SIWE) is proposed. In contrast to prior work, the WER estimators are trained using data that simulates ASR system output. Hypotheses are generated using phonetically similar or linguistically more likely alternative words. In WER estimation experiments, the proposed method reaches a similar performance to ASR system-dependent WER estimators on in-domain data and achieves state-of-the-art performance on out-of-domain data. On the out-of-domain data, the SIWE model outperformed the baseline estimators in root mean square error and Pearson correlation coefficient by relative 17.58% and 18.21%, respectively, on Switchboard and CALLHOME. The performance was further improved when the WER of the training set was close to the WER of the evaluation dataset.
Authors: Fynn Petersen-Frey, Chris Biemann
Abstract: Extracting who says what to whom is a crucial part in analyzing human communication in today's abundance of data such as online news articles. Yet, the lack of annotated data for this task in German news articles severely limits the quality and usability of possible systems. To remedy this, we present a new, freely available, creative-commons-licensed dataset for quotation attribution in German news articles based on WIKINEWS. The dataset provides curated, high-quality annotations across 1000 documents (250,000 tokens) in a fine-grained annotation schema enabling various downstream uses for the dataset. The annotations not only specify who said what but also how, in which context, to whom and define the type of quotation. We specify our annotation schema, describe the creation of the dataset and provide a quantitative analysis. Further, we describe suitable evaluation metrics, apply two existing systems for quotation attribution, discuss their results to evaluate the utility of our dataset and outline use cases of our dataset in downstream tasks.
Authors: Runzhe Zhan, Xinyi Yang, Derek F. Wong, Lidia S. Chao, Yue Zhang
Abstract: While supervised fine-tuning (SFT) has been a straightforward approach for tailoring the output of foundation large language model (LLM) to specific preferences, concerns have been raised about the depth of this alignment, with some critiques suggesting it is merely "superficial". We critically examine this hypothesis within the scope of cross-lingual generation tasks, proposing that the effectiveness of SFT may be constrained by its reliance on prior tokens to guide cross-lingual generation. Based on this crucial insight, and in response to the challenges posed by the costly and limited availability of non-English data for SFT, we introduce a novel training-free alignment method named PreTTY, which employs minimal task-related prior tokens to bridge the foundation LLM and the SFT LLM, achieving comparable performance without training. Experiments on machine translation and part-of-speech tagging across eight languages demonstrate the efficacy of PreTTY in cross-lingual settings. Remarkably, by initiating the decoding process with only one or two prior tokens, foundation LLMs can achieve performance comparable to their SFT counterparts. This method presents a cost-effective alternative to SFT and advances the democratization of multilingual LLMs.
Authors: Jianxiang Zang, Hui Liu
Abstract: Representation-based Siamese networks have risen to popularity in lightweight text matching due to their low deployment and inference costs. While word-level attention mechanisms have been implemented within Siamese networks to improve performance, we propose Feature Attention (FA), a novel downstream block designed to enrich the modeling of dependencies among embedding features. Employing "squeeze-and-excitation" techniques, the FA block dynamically adjusts the emphasis on individual features, enabling the network to concentrate more on features that significantly contribute to the final classification. Building upon FA, we introduce a dynamic "selection" mechanism called Selective Feature Attention (SFA), which leverages a stacked BiGRU Inception structure. The SFA block facilitates multi-scale semantic extraction by traversing different stacked BiGRU layers, encouraging the network to selectively concentrate on semantic information and embedding features across varying levels of abstraction. Both the FA and SFA blocks offer a seamless integration capability with various Siamese networks, showcasing a plug-and-play characteristic. Experimental evaluations conducted across diverse text matching baselines and benchmarks underscore the indispensability of modeling feature attention and the superiority of the "selection" mechanism.
Authors: Tianhui Zhang, Bei Peng, Danushka Bollegala
Abstract: Generative Commonsense Reasoning (GCR) requires a model to reason about a situation using commonsense knowledge, while generating coherent sentences. Although the quality of the generated sentences is crucial, the diversity of the generation is equally important because it reflects the model's ability to use a range of commonsense knowledge facts. Large Language Models (LLMs) have shown proficiency in enhancing the generation quality across various tasks through in-context learning (ICL) using given examples without the need for any fine-tuning. However, the diversity aspect in LLM outputs has not been systematically studied before. To address this, we propose a simple method that diversifies the LLM generations, while preserving their quality. Experimental results on three benchmark GCR datasets show that our method achieves an ideal balance between the quality and diversity. Moreover, the sentences generated by our proposed method can be used as training data to improve diversity in existing commonsense generators.
Authors: Shengnan An, Zexiong Ma, Zeqi Lin, Nanning Zheng, Jian-Guang Lou
Abstract: While many contemporary large language models (LLMs) can process lengthy input, they still struggle to fully utilize information within the long context, known as the lost-in-the-middle challenge. We hypothesize that it stems from insufficient explicit supervision during the long-context training, which fails to emphasize that any position in a long context can hold crucial information. Based on this intuition, our study presents information-intensive (IN2) training, a purely data-driven solution to overcome lost-in-the-middle. Specifically, IN2 training leverages a synthesized long-context question-answer dataset, where the answer requires (1) fine-grained information awareness on a short segment (~128 tokens) within a synthesized long context (4K-32K tokens), and (2) the integration and reasoning of information from two or more short segments. Through applying this information-intensive training on Mistral-7B, we present FILM-7B (FILl-in-the-Middle). To thoroughly assess the ability of FILM-7B for utilizing long contexts, we design three probing tasks that encompass various context styles (document, code, and structured-data context) and information retrieval patterns (forward, backward, and bi-directional retrieval). The probing results demonstrate that FILM-7B can robustly retrieve information from different positions in its 32K context window. Beyond these probing tasks, FILM-7B significantly improves the performance on real-world long-context tasks (e.g., 23.5->26.9 F1 score on NarrativeQA), while maintaining a comparable performance on short-context tasks (e.g., 59.3->59.2 accuracy on MMLU). Github Link: https://github.com/microsoft/FILM.
Authors: Harman Singh, Nitish Gupta, Shikhar Bharadwaj, Dinesh Tewari, Partha Talukdar
Abstract: As large language models (LLMs) see increasing adoption across the globe, it is imperative for LLMs to be representative of the linguistic diversity of the world. India is a linguistically diverse country of 1.4 Billion people. To facilitate research on multilingual LLM evaluation, we release IndicGenBench - the largest benchmark for evaluating LLMs on user-facing generation tasks across a diverse set 29 of Indic languages covering 13 scripts and 4 language families. IndicGenBench is composed of diverse generation tasks like cross-lingual summarization, machine translation, and cross-lingual question answering. IndicGenBench extends existing benchmarks to many Indic languages through human curation providing multi-way parallel evaluation data for many under-represented Indic languages for the first time. We evaluate a wide range of proprietary and open-source LLMs including GPT-3.5, GPT-4, PaLM-2, mT5, Gemma, BLOOM and LLaMA on IndicGenBench in a variety of settings. The largest PaLM-2 models performs the best on most tasks, however, there is a significant performance gap in all languages compared to English showing that further research is needed for the development of more inclusive multilingual language models. IndicGenBench is released at www.github.com/google-research-datasets/indic-gen-bench
Authors: Derek Jacoby, Tianyi Zhang, Aanchan Mohan, Yvonne Coady
Abstract: A problem with many current Large Language Model (LLM) driven spoken dialogues is the response time. Some efforts such as Groq address this issue by lightning fast processing of the LLM, but we know from the cognitive psychology literature that in human-to-human dialogue often responses occur prior to the speaker completing their utterance. No amount of delay for LLM processing is acceptable if we wish to maintain human dialogue latencies. In this paper, we discuss methods for understanding an utterance in close to real time and generating a response so that the system can comply with human-level conversational turn delays. This means that the information content of the final part of the speaker's utterance is lost to the LLM. Using the Google NaturalQuestions (NQ) database, our results show GPT-4 can effectively fill in missing context from a dropped word at the end of a question over 60% of the time. We also provide some examples of utterances and the impacts of this information loss on the quality of LLM response in the context of an avatar that is currently under development. These results indicate that a simple classifier could be used to determine whether a question is semantically complete, or requires a filler phrase to allow a response to be generated within human dialogue time constraints.
Authors: Yifan Jiang, Filip Ilievski, Kaixin Ma
Abstract: While vertical thinking relies on logical and commonsense reasoning, lateral thinking requires systems to defy commonsense associations and overwrite them through unconventional thinking. Lateral thinking has been shown to be challenging for current models but has received little attention. A recent benchmark, BRAINTEASER, aims to evaluate current models' lateral thinking ability in a zero-shot setting. In this paper, we split the original benchmark to also support fine-tuning setting and present SemEval Task 9: BRAIN-TEASER(S), the first task at this competition designed to test the system's reasoning and lateral thinking ability. As a popular task, BRAINTEASER(S)'s two subtasks receive 483 team submissions from 182 participants during the competition. This paper provides a fine-grained system analysis of the competition results, together with a reflection on what this means for the ability of the systems to reason laterally. We hope that the BRAINTEASER(S) subtasks and findings in this paper can stimulate future work on lateral thinking and robust reasoning by computational models.
Authors: Xiang Tao, Liang Wang, Qiang Liu, Shu Wu, Liang Wang
Abstract: Due to the rapid spread of rumors on social media, rumor detection has become an extremely important challenge. Recently, numerous rumor detection models which utilize textual information and the propagation structure of events have been proposed. However, these methods overlook the importance of semantic evolvement information of event in propagation process, which is often challenging to be truly learned in supervised training paradigms and traditional rumor detection methods. To address this issue, we propose a novel semantic evolvement enhanced Graph Autoencoder for Rumor Detection (GARD) model in this paper. The model learns semantic evolvement information of events by capturing local semantic changes and global semantic evolvement information through specific graph autoencoder and reconstruction strategies. By combining semantic evolvement information and propagation structure information, the model achieves a comprehensive understanding of event propagation and perform accurate and robust detection, while also detecting rumors earlier by capturing semantic evolvement information in the early stages. Moreover, in order to enhance the model's ability to learn the distinct patterns of rumors and non-rumors, we introduce a uniformity regularizer to further improve the model's performance. Experimental results on three public benchmark datasets confirm the superiority of our GARD method over the state-of-the-art approaches in both overall performance and early rumor detection.
Authors: Albert Rilliard, David Doukhan, R\'emi Uro, Simon Devauchelle
Abstract: We present a diachronic acoustic analysis of the voice of 1023 speakers from French media archives. The speakers are spread across 32 categories based on four periods (years 1955/56, 1975/76, 1995/96, 2015/16), four age groups (20-35; 36-50; 51-65, >65), and two genders. The fundamental frequency ($F_0$) and the first four formants (F1-4) were estimated. Procedures used to ensure the quality of these estimations on heterogeneous data are described. From each speaker's $F_0$ distribution, the base-$F_0$ value was calculated to estimate the register. Average vocal tract length was estimated from formant frequencies. Base-$F_0$ and vocal tract length were fit by linear mixed models to evaluate how they may have changed across time periods and genders, corrected for age effects. Results show an effect of the period with a tendency to lower voices, independently of gender. A lowering of pitch is observed with age for female but not male speakers.
Authors: Eric Slyman, Stefan Lee, Scott Cohen, Kushal Kafle
Abstract: Recent dataset deduplication techniques have demonstrated that content-aware dataset pruning can dramatically reduce the cost of training Vision-Language Pretrained (VLP) models without significant performance losses compared to training on the original dataset. These results have been based on pruning commonly used image-caption datasets collected from the web -- datasets that are known to harbor harmful social biases that may then be codified in trained models. In this work, we evaluate how deduplication affects the prevalence of these biases in the resulting trained models and introduce an easy-to-implement modification to the recent SemDeDup algorithm that can reduce the negative effects that we observe. When examining CLIP-style models trained on deduplicated variants of LAION-400M, we find our proposed FairDeDup algorithm consistently leads to improved fairness metrics over SemDeDup on the FairFace and FACET datasets while maintaining zero-shot performance on CLIP benchmarks.
Authors: Sakher Khalil Alqaaidi, Krzysztof Kochut
Abstract: Knowledge Graphs (KGs) are widely employed in artificial intelligence applications, such as question-answering and recommendation systems. However, KGs are frequently found to be incomplete. While much of the existing literature focuses on predicting missing nodes for given incomplete KG triples, there remains an opportunity to complete KGs by exploring relations between existing nodes, a task known as relation prediction. In this study, we propose a relations prediction model that harnesses both textual and structural information within KGs. Our approach integrates walks-based embeddings with language model embeddings to effectively represent nodes. We demonstrate that our model achieves competitive results in the relation prediction task when evaluated on a widely used dataset.
Authors: Divyansh Agarwal, Alexander R. Fabbri, Philippe Laban, Shafiq Joty, Caiming Xiong, Chien-Sheng Wu
Abstract: Prompt leakage in large language models (LLMs) poses a significant security and privacy threat, particularly in retrieval-augmented generation (RAG) systems. However, leakage in multi-turn LLM interactions along with mitigation strategies has not been studied in a standardized manner. This paper investigates LLM vulnerabilities against prompt leakage across 4 diverse domains and 10 closed- and open-source LLMs. Our unique multi-turn threat model leverages the LLM's sycophancy effect and our analysis dissects task instruction and knowledge leakage in the LLM response. In a multi-turn setting, our threat model elevates the average attack success rate (ASR) to 86.2%, including a 99% leakage with GPT-4 and claude-1.3. We find that some black-box LLMs like Gemini show variable susceptibility to leakage across domains - they are more likely to leak contextual knowledge in the news domain compared to the medical domain. Our experiments measure specific effects of 6 black-box defense strategies, including a query-rewriter in the RAG scenario. Our proposed multi-tier combination of defenses still has an ASR of 5.3% for black-box LLMs, indicating room for enhancement and future direction for LLM security research.
Authors: An Yan, Zhengyuan Yang, Junda Wu, Wanrong Zhu, Jianwei Yang, Linjie Li, Kevin Lin, Jianfeng Wang, Julian McAuley, Jianfeng Gao, Lijuan Wang
Abstract: Set-of-Mark (SoM) Prompting unleashes the visual grounding capability of GPT-4V, by enabling the model to associate visual objects with tags inserted on the image. These tags, marked with alphanumerics, can be indexed via text tokens for easy reference. Despite the extraordinary performance from GPT-4V, we observe that other Multimodal Large Language Models (MLLMs) struggle to understand these visual tags. To promote the learning of SoM prompting for open-source models, we propose a new learning paradigm: "list items one by one," which asks the model to enumerate and describe all visual tags placed on the image following the alphanumeric orders of tags. By integrating our curated dataset with other visual instruction tuning datasets, we are able to equip existing MLLMs with the SoM prompting ability. Furthermore, we evaluate our finetuned SoM models on five MLLM benchmarks. We find that this new dataset, even in a relatively small size (10k-30k images with tags), significantly enhances visual reasoning capabilities and reduces hallucinations for MLLMs. Perhaps surprisingly, these improvements persist even when the visual tags are omitted from input images during inference. This suggests the potential of "list items one by one" as a new paradigm for training MLLMs, which strengthens the object-text alignment through the use of visual tags in the training stage. Finally, we conduct analyses by probing trained models to understand the working mechanism of SoM. Our code and data are available at \url{https://github.com/zzxslp/SoM-LLaVA}.
Authors: Emre Can Acikgoz, Osman Batur \.Ince, Rayene Bench, Arda An{\i}l Boz, \.Ilker Kesen, Aykut Erdem, Erkut Erdem
Abstract: The integration of Large Language Models (LLMs) into healthcare promises to transform medical diagnostics, research, and patient care. Yet, the progression of medical LLMs faces obstacles such as complex training requirements, rigorous evaluation demands, and the dominance of proprietary models that restrict academic exploration. Transparent, comprehensive access to LLM resources is essential for advancing the field, fostering reproducibility, and encouraging innovation in healthcare AI. We present Hippocrates, an open-source LLM framework specifically developed for the medical domain. In stark contrast to previous efforts, it offers unrestricted access to its training datasets, codebase, checkpoints, and evaluation protocols. This open approach is designed to stimulate collaborative research, allowing the community to build upon, refine, and rigorously evaluate medical LLMs within a transparent ecosystem. Also, we introduce Hippo, a family of 7B models tailored for the medical domain, fine-tuned from Mistral and LLaMA2 through continual pre-training, instruction tuning, and reinforcement learning from human and AI feedback. Our models outperform existing open medical LLMs models by a large-margin, even surpassing models with 70B parameters. Through Hippocrates, we aspire to unlock the full potential of LLMs not just to advance medical knowledge and patient care but also to democratize the benefits of AI research in healthcare, making them available across the globe.
Authors: Zhaolin Gao, Jonathan D. Chang, Wenhao Zhan, Owen Oertell, Gokul Swamy, Kiant\'e Brantley, Thorsten Joachims, J. Andrew Bagnell, Jason D. Lee, Wen Sun
Abstract: While originally developed for continuous control problems, Proximal Policy Optimization (PPO) has emerged as the work-horse of a variety of reinforcement learning (RL) applications including the fine-tuning of generative models. Unfortunately, PPO requires multiple heuristics to enable stable convergence (e.g. value networks, clipping) and is notorious for its sensitivity to the precise implementation of these components. In response, we take a step back and ask what a minimalist RL algorithm for the era of generative models would look like. We propose REBEL, an algorithm that cleanly reduces the problem of policy optimization to regressing the relative rewards via a direct policy parameterization between two completions to a prompt, enabling strikingly lightweight implementation. In theory, we prove that fundamental RL algorithms like Natural Policy Gradient can be seen as variants of REBEL, which allows us to match the strongest known theoretical guarantees in terms of convergence and sample complexity in the RL literature. REBEL can also cleanly incorporate offline data and handle the intransitive preferences we frequently see in practice. Empirically, we find that REBEL provides a unified approach to language modeling and image generation with stronger or similar performance as PPO and DPO, all while being simpler to implement and more computationally tractable than PPO.
Authors: Haizhou Shi, Zihao Xu, Hengyi Wang, Weiyi Qin, Wenyuan Wang, Yibin Wang, Hao Wang
Abstract: The recent success of large language models (LLMs) trained on static, pre-collected, general datasets has sparked numerous research directions and applications. One such direction addresses the non-trivial challenge of integrating pre-trained LLMs into dynamic data distributions, task structures, and user preferences. Pre-trained LLMs, when tailored for specific needs, often experience significant performance degradation in previous knowledge domains -- a phenomenon known as "catastrophic forgetting". While extensively studied in the continual learning (CL) community, it presents new manifestations in the realm of LLMs. In this survey, we provide a comprehensive overview of the current research progress on LLMs within the context of CL. This survey is structured into four main sections: we first describe an overview of continually learning LLMs, consisting of two directions of continuity: vertical continuity (or vertical continual learning), i.e., continual adaptation from general to specific capabilities, and horizontal continuity (or horizontal continual learning), i.e., continual adaptation across time and domains (Section 3). We then summarize three stages of learning LLMs in the context of modern CL: Continual Pre-Training (CPT), Domain-Adaptive Pre-training (DAP), and Continual Fine-Tuning (CFT) (Section 4). Then we provide an overview of evaluation protocols for continual learning with LLMs, along with the current available data sources (Section 5). Finally, we discuss intriguing questions pertaining to continual learning for LLMs (Section 6). The full list of papers examined in this survey is available at https://github.com/Wang-ML-Lab/llm-continual-learning-survey.
URLs: https://github.com/Wang-ML-Lab/llm-continual-learning-survey.
Authors: Chujie Zheng, Ziqi Wang, Heng Ji, Minlie Huang, Nanyun Peng
Abstract: Although the capabilities of large language models (LLMs) ideally scale up with increasing data and compute, they are inevitably constrained by limited resources in reality. Suppose we have a moderately trained LLM (e.g., trained to align with human preference) in hand, can we further exploit its potential and cheaply acquire a stronger model? In this paper, we propose a simple method called ExPO to boost LLMs' alignment with human preference. ExPO assumes that a medium-aligned model can be interpolated between a less-aligned (weaker) model, e.g., the initial SFT model, and a better-aligned (stronger) one, thereby directly obtaining this stronger model by extrapolating from the weights of the former two relatively weaker models. On the AlpacaEval 2.0 benchmark, we show that ExPO pushes models trained with less preference data (e.g., 10% or 20%) to reach and even surpass the fully-trained one, without any additional training. Furthermore, ExPO also significantly improves off-the-shelf DPO/RLHF models and exhibits decent scalability across model sizes from 7B to 70B. Our work demonstrates the efficacy of model extrapolation in exploiting LLMs' capabilities, suggesting a promising direction that deserves future exploration.
Authors: Ye Fang, Zeyi Sun, Tong Wu, Jiaqi Wang, Ziwei Liu, Gordon Wetzstein, Dahua Lin
Abstract: Physically realistic materials are pivotal in augmenting the realism of 3D assets across various applications and lighting conditions. However, existing 3D assets and generative models often lack authentic material properties. Manual assignment of materials using graphic software is a tedious and time-consuming task. In this paper, we exploit advancements in Multimodal Large Language Models (MLLMs), particularly GPT-4V, to present a novel approach, Make-it-Real: 1) We demonstrate that GPT-4V can effectively recognize and describe materials, allowing the construction of a detailed material library. 2) Utilizing a combination of visual cues and hierarchical text prompts, GPT-4V precisely identifies and aligns materials with the corresponding components of 3D objects. 3) The correctly matched materials are then meticulously applied as reference for the new SVBRDF material generation according to the original diffuse map, significantly enhancing their visual authenticity. Make-it-Real offers a streamlined integration into the 3D content creation workflow, showcasing its utility as an essential tool for developers of 3D assets.
Authors: Jacob Eisenstein, Daniel Andor, Bernd Bohnet, Michael Collins, David Mimno
Abstract: Explainable question answering systems should produce not only accurate answers but also rationales that justify their reasoning and allow humans to check their work. But what sorts of rationales are useful and how can we train systems to produce them? We propose a new style of rationale for open-book question answering, called \emph{markup-and-mask}, which combines aspects of extractive and free-text explanations. In the markup phase, the passage is augmented with free-text markup that enables each sentence to stand on its own outside the discourse context. In the masking phase, a sub-span of the marked-up passage is selected. To train a system to produce markup-and-mask rationales without annotations, we leverage in-context learning. Specifically, we generate silver annotated data by sending a series of prompts to a frozen pretrained language model, which acts as a teacher. We then fine-tune a smaller student model by training on the subset of rationales that led to correct answers. The student is "honest" in the sense that it is a pipeline: the rationale acts as a bottleneck between the passage and the answer, while the "untrusted" teacher operates under no such constraints. Thus, we offer a new way to build trustworthy pipeline systems from a combination of end-task annotations and frozen pretrained language models.
Authors: Shauli Ravfogel, Valentina Pyatkin, Amir DN Cohen, Avshalom Manevich, Yoav Goldberg
Abstract: Identifying texts with a given semantics is central for many information seeking scenarios. Similarity search over vector embeddings appear to be central to this ability, yet the similarity reflected in current text embeddings is corpus-driven, and is inconsistent and sub-optimal for many use cases. What, then, is a good notion of similarity for effective retrieval of text? We identify the need to search for texts based on abstract descriptions of their content, and the corresponding notion of \emph{description based similarity}. We demonstrate the inadequacy of current text embeddings and propose an alternative model that significantly improves when used in standard nearest neighbor search. The model is trained using positive and negative pairs sourced through prompting a LLM, demonstrating how data from LLMs can be used for creating new capabilities not immediately possible using the original model.
Authors: Angus R. Williams, Hannah Rose Kirk, Liam Burke, Yi-Ling Chung, Ivan Debono, Pica Johansson, Francesca Stevens, Jonathan Bright, Scott A. Hale
Abstract: Public figures receive a disproportionate amount of abuse on social media, impacting their active participation in public life. Automated systems can identify abuse at scale but labelling training data is expensive, complex and potentially harmful. So, it is desirable that systems are efficient and generalisable, handling both shared and specific aspects of online abuse. We explore the dynamics of cross-group text classification in order to understand how well classifiers trained on one domain or demographic can transfer to others, with a view to building more generalisable abuse classifiers. We fine-tune language models to classify tweets targeted at public figures across DOmains (sport and politics) and DemOgraphics (women and men) using our novel DODO dataset, containing 28,000 labelled entries, split equally across four domain-demographic pairs. We find that (i) small amounts of diverse data are hugely beneficial to generalisation and model adaptation; (ii) models transfer more easily across demographics but models trained on cross-domain data are more generalisable; (iii) some groups contribute more to generalisability than others; and (iv) dataset similarity is a signal of transferability.
Authors: Kaihang Pan, Juncheng Li, Wenjie Wang, Hao Fei, Hongye Song, Wei Ji, Jun Lin, Xiaozhong Liu, Tat-Seng Chua, Siliang Tang
Abstract: Recent studies indicate that dense retrieval models struggle to perform well on a wide variety of retrieval tasks that lack dedicated training data, as different retrieval tasks often entail distinct search intents. To address this challenge, in this work we leverage instructions to flexibly describe retrieval intents and introduce I3, a unified retrieval system that performs Intent-Introspective retrieval across various tasks, conditioned on Instructions without any task-specific training. I3 innovatively incorporates a pluggable introspector in a parameter-isolated manner to comprehend specific retrieval intents by jointly reasoning over the input query and instruction, and seamlessly integrates the introspected intent into the original retrieval model for intent-aware retrieval. Furthermore, we propose progressively-pruned intent learning. It utilizes extensive LLM-generated data to train I3 phase-by-phase, embodying two key designs: progressive structure pruning and drawback extrapolation-based data refinement. Extensive experiments show that in the BEIR benchmark, I3 significantly outperforms baseline methods designed with task-specific retrievers, achieving state-of-the-art zero-shot performance without any task-specific tuning.
Authors: Belen Alastruey, Aleix Sant, Gerard I. G\'allego, David Dale, Marta R. Costa-juss\`a
Abstract: Speech-to-Speech and Speech-to-Text translation are currently dynamic areas of research. In our commitment to advance these fields, we present SpeechAlign, a framework designed to evaluate the underexplored field of source-target alignment in speech models. The SpeechAlign framework has two core components. First, to tackle the absence of suitable evaluation datasets, we introduce the Speech Gold Alignment dataset, built upon a English-German text translation gold alignment dataset. Secondly, we introduce two novel metrics, Speech Alignment Error Rate (SAER) and Time-weighted Speech Alignment Error Rate (TW-SAER), which enable the evaluation of alignment quality within speech models. While the former gives equal importance to each word, the latter assigns weights based on the length of the words in the speech signal. By publishing SpeechAlign we provide an accessible evaluation framework for model assessment, and we employ it to benchmark open-source Speech Translation models. In doing so, we contribute to the ongoing research progress within the fields of Speech-to-Speech and Speech-to-Text translation.
Authors: In Gim, Guojun Chen, Seung-seob Lee, Nikhil Sarda, Anurag Khandelwal, Lin Zhong
Abstract: We present Prompt Cache, an approach for accelerating inference for large language models (LLM) by reusing attention states across different LLM prompts. Many input prompts have overlapping text segments, such as system messages, prompt templates, and documents provided for context. Our key insight is that by precomputing and storing the attention states of these frequently occurring text segments on the inference server, we can efficiently reuse them when these segments appear in user prompts. Prompt Cache employs a schema to explicitly define such reusable text segments, called prompt modules. The schema ensures positional accuracy during attention state reuse and provides users with an interface to access cached states in their prompt. Using a prototype implementation, we evaluate Prompt Cache across several LLMs. We show that Prompt Cache significantly reduce latency in time-to-first-token, especially for longer prompts such as document-based question answering and recommendations. The improvements range from 8x for GPU-based inference to 60x for CPU-based inference, all while maintaining output accuracy and without the need for model parameter modifications.
Authors: Aakanksha Naik, Bailey Kuehl, Erin Bransom, Doug Downey, Tom Hope
Abstract: Extracting fine-grained experimental findings from literature can provide dramatic utility for scientific applications. Prior work has developed annotation schemas and datasets for limited aspects of this problem, failing to capture the real-world complexity and nuance required. Focusing on biomedicine, this work presents CARE -- a new IE dataset for the task of extracting clinical findings. We develop a new annotation schema capturing fine-grained findings as n-ary relations between entities and attributes, which unifies phenomena challenging for current IE systems such as discontinuous entity spans, nested relations, variable arity n-ary relations and numeric results in a single schema. We collect extensive annotations for 700 abstracts from two sources: clinical trials and case reports. We also demonstrate the generalizability of our schema to the computer science and materials science domains. We benchmark state-of-the-art IE systems on CARE, showing that even models such as GPT4 struggle. We release our resources to advance research on extracting and aggregating literature findings.
Authors: Fanghua Ye, Mingming Yang, Jianhui Pang, Longyue Wang, Derek F. Wong, Emine Yilmaz, Shuming Shi, Zhaopeng Tu
Abstract: The proliferation of open-source Large Language Models (LLMs) from various institutions has highlighted the urgent need for comprehensive evaluation methods. However, current evaluation platforms, such as the widely recognized HuggingFace open LLM leaderboard, neglect a crucial aspect -- uncertainty, which is vital for thoroughly assessing LLMs. To bridge this gap, we introduce a new benchmarking approach for LLMs that integrates uncertainty quantification. Our examination involves eight LLMs (LLM series) spanning five representative natural language processing tasks. Our findings reveal that: I) LLMs with higher accuracy may exhibit lower certainty; II) Larger-scale LLMs may display greater uncertainty compared to their smaller counterparts; and III) Instruction-finetuning tends to increase the uncertainty of LLMs. These results underscore the significance of incorporating uncertainty in the evaluation of LLMs.
Authors: Berk Atil, Mahsa Sheikhi Karizaki, Rebecca J. Passonneau
Abstract: With an increasing focus in STEM education on critical thinking skills, science writing plays an ever more important role in curricula that stress inquiry skills. A recently published dataset of two sets of college level lab reports from an inquiry-based physics curriculum relies on analytic assessment rubrics that utilize multiple dimensions, specifying subject matter knowledge and general components of good explanations. Each analytic dimension is assessed on a 6-point scale, to provide detailed feedback to students that can help them improve their science writing skills. Manual assessment can be slow, and difficult to calibrate for consistency across all students in large classes. While much work exists on automated assessment of open-ended questions in STEM subjects, there has been far less work on long-form writing such as lab reports. We present an end-to-end neural architecture that has separate verifier and assessment modules, inspired by approaches to Open Domain Question Answering (OpenQA). VerAs first verifies whether a report contains any content relevant to a given rubric dimension, and if so, assesses the relevant sentences. On the lab reports, VerAs outperforms multiple baselines based on OpenQA systems or Automated Essay Scoring (AES). VerAs also performs well on an analytic rubric for middle school physics essays.
Authors: Sankalan Pal Chowdhury, Vil\'em Zouhar, Mrinmaya Sachan
Abstract: Large Language Models (LLMs) have found several use cases in education, ranging from automatic question generation to essay evaluation. In this paper, we explore the potential of using Large Language Models (LLMs) to author Intelligent Tutoring Systems. A common pitfall of LLMs is their straying from desired pedagogical strategies such as leaking the answer to the student, and in general, providing no guarantees. We posit that while LLMs with certain guardrails can take the place of subject experts, the overall pedagogical design still needs to be handcrafted for the best learning results. Based on this principle, we create a sample end-to-end tutoring system named MWPTutor, which uses LLMs to fill in the state space of a pre-defined finite state transducer. This approach retains the structure and the pedagogy of traditional tutoring systems that has been developed over the years by learning scientists but brings in additional flexibility of LLM-based approaches. Through a human evaluation study on two datasets based on math word problems, we show that our hybrid approach achieves a better overall tutoring score than an instructed, but otherwise free-form, GPT-4. MWPTutor is completely modular and opens up the scope for the community to improve its performance by improving individual modules or using different teaching strategies that it can follow.
Authors: Xiaolong Wang, Yile Wang, Sijie Cheng, Peng Li, Yang Liu
Abstract: Recent work has made a preliminary attempt to use large language models (LLMs) to solve the stance detection task, showing promising results. However, considering that stance detection usually requires detailed background knowledge, the vanilla reasoning method may neglect the domain knowledge to make a professional and accurate analysis. Thus, there is still room for improvement of LLMs reasoning, especially in leveraging the generation capability of LLMs to simulate specific experts (i.e., multi-agents) to detect the stance. In this paper, different from existing multi-agent works that require detailed descriptions and use fixed experts, we propose a Dynamic Experienced Expert Modeling (DEEM) method which can leverage the generated experienced experts and let LLMs reason in a semi-parametric way, making the experts more generalizable and reliable. Experimental results demonstrate that DEEM consistently achieves the best results on three standard benchmarks, outperforms methods with self-consistency reasoning, and reduces the bias of LLMs.
Authors: Bing Xue, Charles Alba, Joanna Abraham, Thomas Kannampallil, Chenyang Lu
Abstract: Predicting postoperative risk can inform effective care management & planning. We explored large language models (LLMs) in predicting postoperative risk through clinical texts using various tuning strategies. Records spanning 84,875 patients from Barnes Jewish Hospital (BJH) between 2018 & 2021, with a mean duration of follow-up based on the length of postoperative ICU stay less than 7 days, were utilized. Methods were replicated on the MIMIC-III dataset. Outcomes included 30-day mortality, pulmonary embolism (PE) & pneumonia. Three domain adaptation & finetuning strategies were implemented for three LLMs (BioGPT, ClinicalBERT & BioClinicalBERT): self-supervised objectives; incorporating labels with semi-supervised fine-tuning; & foundational modelling through multi-task learning. Model performance was compared using the AUROC & AUPRC for classification tasks & MSE & R2 for regression tasks. Cohort had a mean age of 56.9 (sd: 16.8) years; 50.3% male; 74% White. Pre-trained LLMs outperformed traditional word embeddings, with absolute maximal gains of 38.3% for AUROC & 14% for AUPRC. Adapting models through self-supervised finetuning further improved performance by 3.2% for AUROC & 1.5% for AUPRC Incorporating labels into the finetuning procedure further boosted performances, with semi-supervised finetuning improving by 1.8% for AUROC & 2% for AUPRC & foundational modelling improving by 3.6% for AUROC & 2.6% for AUPRC compared to self-supervised finetuning. Pre-trained clinical LLMs offer opportunities for postoperative risk predictions with unseen data, & further improvements from finetuning suggests benefits in adapting pre-trained models to note-specific perioperative use cases. Incorporating labels can further boost performance. The superior performance of foundational models suggests the potential of task-agnostic learning towards the generalizable LLMs in perioperative care.
Authors: Miaomiao Li, Jiaqi Zhu, Yang Wang, Yi Yang, Yilin Li, Hongan Wang
Abstract: Weakly supervised text classification (WSTC), also called zero-shot or dataless text classification, has attracted increasing attention due to its applicability in classifying a mass of texts within the dynamic and open Web environment, since it requires only a limited set of seed words (label names) for each category instead of labeled data. With the help of recently popular prompting Pre-trained Language Models (PLMs), many studies leveraged manually crafted and/or automatically identified verbalizers to estimate the likelihood of categories, but they failed to differentiate the effects of these category-indicative words, let alone capture their correlations and realize adaptive adjustments according to the unlabeled corpus. In this paper, in order to let the PLM effectively understand each category, we at first propose a novel form of rule-based knowledge using logical expressions to characterize the meanings of categories. Then, we develop a prompting PLM-based approach named RulePrompt for the WSTC task, consisting of a rule mining module and a rule-enhanced pseudo label generation module, plus a self-supervised fine-tuning module to make the PLM align with this task. Within this framework, the inaccurate pseudo labels assigned to texts and the imprecise logical rules associated with categories mutually enhance each other in an alternative manner. That establishes a self-iterative closed loop of knowledge (rule) acquisition and utilization, with seed words serving as the starting point. Extensive experiments validate the effectiveness and robustness of our approach, which markedly outperforms state-of-the-art weakly supervised methods. What is more, our approach yields interpretable category rules, proving its advantage in disambiguating easily-confused categories.
Authors: Gemini Team, Machel Reid (Dima), Nikolay Savinov (Dima), Denis Teplyashin (Dima), Dmitry (Dima), Lepikhin, Timothy Lillicrap, Jean-baptiste Alayrac, Radu Soricut, Angeliki Lazaridou, Orhan Firat, Julian Schrittwieser, Ioannis Antonoglou, Rohan Anil, Sebastian Borgeaud, Andrew Dai, Katie Millican, Ethan Dyer, Mia Glaese, Thibault Sottiaux, Benjamin Lee, Fabio Viola, Malcolm Reynolds, Yuanzhong Xu, James Molloy, Jilin Chen, Michael Isard, Paul Barham, Tom Hennigan, Ross McIlroy, Melvin Johnson, Johan Schalkwyk, Eli Collins, Eliza Rutherford, Erica Moreira, Kareem Ayoub, Megha Goel, Clemens Meyer, Gregory Thornton, Zhen Yang, Henryk Michalewski, Zaheer Abbas, Nathan Schucher, Ankesh Anand, Richard Ives, James Keeling, Karel Lenc, Salem Haykal, Siamak Shakeri, Pranav Shyam, Aakanksha Chowdhery, Roman Ring, Stephen Spencer, Eren Sezener, Luke Vilnis, Oscar Chang, Nobuyuki Morioka, George Tucker, Ce Zheng, Oliver Woodman, Nithya Attaluri, Tomas Kocisky, Evgenii Eltyshev, Xi Chen, Timothy Chung, Vittorio Selo, Siddhartha Brahma, Petko Georgiev, Ambrose Slone, Zhenkai Zhu, James Lottes, Siyuan Qiao, Ben Caine, Sebastian Riedel, Alex Tomala, Martin Chadwick, Juliette Love, Peter Choy, Sid Mittal, Neil Houlsby, Yunhao Tang, Matthew Lamm, Libin Bai, Qiao Zhang, Luheng He, Yong Cheng, Peter Humphreys, Yujia Li, Sergey Brin, Albin Cassirer, Yingjie Miao, Lukas Zilka, Taylor Tobin, Kelvin Xu, Lev Proleev, Daniel Sohn, Alberto Magni, Lisa Anne Hendricks, Isabel Gao, Santiago Ontanon, Oskar Bunyan, Nathan Byrd, Abhanshu Sharma, Biao Zhang, Mario Pinto, Rishika Sinha, Harsh Mehta, Dawei Jia, Sergi Caelles, Albert Webson, Alex Morris, Becca Roelofs, Yifan Ding, Robin Strudel, Xuehan Xiong, Marvin Ritter, Mostafa Dehghani, Rahma Chaabouni, Abhijit Karmarkar, Guangda Lai, Fabian Mentzer, Bibo Xu, YaGuang Li, Yujing Zhang, Tom Le Paine, Alex Goldin, Behnam Neyshabur, Kate Baumli, Anselm Levskaya, Michael Laskin, Wenhao Jia, Jack W. Rae, Kefan Xiao, Antoine He, Skye Giordano, Lakshman Yagati, Jean-Baptiste Lespiau, Paul Natsev, Sanjay Ganapathy, Fangyu Liu, Danilo Martins, Nanxin Chen, Yunhan Xu, Megan Barnes, Rhys May, Arpi Vezer, Junhyuk Oh, Ken Franko, Sophie Bridgers, Ruizhe Zhao, Boxi Wu, Basil Mustafa, Sean Sechrist, Emilio Parisotto, Thanumalayan Sankaranarayana Pillai, Chris Larkin, Chenjie Gu, Christina Sorokin, Maxim Krikun, Alexey Guseynov, Jessica Landon, Romina Datta, Alexander Pritzel, Phoebe Thacker, Fan Yang, Kevin Hui, Anja Hauth, Chih-Kuan Yeh, David Barker, Justin Mao-Jones, Sophia Austin, Hannah Sheahan, Parker Schuh, James Svensson, Rohan Jain, Vinay Ramasesh, Anton Briukhov, Da-Woon Chung, Tamara von Glehn, Christina Butterfield, Priya Jhakra, Matthew Wiethoff, Justin Frye, Jordan Grimstad, Beer Changpinyo, Charline Le Lan, Anna Bortsova, Yonghui Wu, Paul Voigtlaender, Tara Sainath, Shane Gu, Charlotte Smith, Will Hawkins, Kris Cao, James Besley, Srivatsan Srinivasan, Mark Omernick, Colin Gaffney, Gabriela Surita, Ryan Burnell, Bogdan Damoc, Junwhan Ahn, Andrew Brock, Mantas Pajarskas, Anastasia Petrushkina, Seb Noury, Lorenzo Blanco, Kevin Swersky, Arun Ahuja, Thi Avrahami, Vedant Misra, Raoul de Liedekerke, Mariko Iinuma, Alex Polozov, Sarah York, George van den Driessche, Paul Michel, Justin Chiu, Rory Blevins, Zach Gleicher, Adri\`a Recasens, Alban Rrustemi, Elena Gribovskaya, Aurko Roy, Wiktor Gworek, S\'ebastien M. R. Arnold, Lisa Lee, James Lee-Thorp, Marcello Maggioni, Enrique Piqueras, Kartikeya Badola, Sharad Vikram, Lucas Gonzalez, Anirudh Baddepudi, Evan Senter, Jacob Devlin, James Qin, Michael Azzam, Maja Trebacz, Martin Polacek, Kashyap Krishnakumar, Shuo-yiin Chang, Matthew Tung, Ivo Penchev, Rishabh Joshi, Kate Olszewska, Carrie Muir, Mateo Wirth, Ale Jakse Hartman, Josh Newlan, Sheleem Kashem, Vijay Bolina, Elahe Dabir, Joost van Amersfoort, Zafarali Ahmed, James Cobon-Kerr, Aishwarya Kamath, Arnar Mar Hrafnkelsson, Le Hou, Ian Mackinnon, Alexandre Frechette, Eric Noland, Xiance Si, Emanuel Taropa, Dong Li, Phil Crone, Anmol Gulati, S\'ebastien Cevey, Jonas Adler, Ada Ma, David Silver, Simon Tokumine, Richard Powell, Stephan Lee, Kiran Vodrahalli, Samer Hassan, Diana Mincu, Antoine Yang, Nir Levine, Jenny Brennan, Mingqiu Wang, Sarah Hodkinson, Jeffrey Zhao, Josh Lipschultz, Aedan Pope, Michael B. Chang, Cheng Li, Laurent El Shafey, Michela Paganini, Sholto Douglas, Bernd Bohnet, Fabio Pardo, Seth Odoom, Mihaela Rosca, Cicero Nogueira dos Santos, Kedar Soparkar, Arthur Guez, Tom Hudson, Steven Hansen, Chulayuth Asawaroengchai, Ravi Addanki, Tianhe Yu, Wojciech Stokowiec, Mina Khan, Justin Gilmer, Jaehoon Lee, Carrie Grimes Bostock, Keran Rong, Jonathan Caton, Pedram Pejman, Filip Pavetic, Geoff Brown, Vivek Sharma, Mario Lu\v{c}i\'c, Rajkumar Samuel, Josip Djolonga, Amol Mandhane, Lars Lowe Sj\"osund, Elena Buchatskaya, Elspeth White, Natalie Clay, Jiepu Jiang, Hyeontaek Lim, Ross Hemsley, Zeyncep Cankara, Jane Labanowski, Nicola De Cao, David Steiner, Sayed Hadi Hashemi, Jacob Austin, Anita Gergely, Tim Blyth, Joe Stanton, Kaushik Shivakumar, Aditya Siddhant, Anders Andreassen, Carlos Araya, Nikhil Sethi, Rakesh Shivanna, Steven Hand, Ankur Bapna, Ali Khodaei, Antoine Miech, Garrett Tanzer, Andy Swing, Shantanu Thakoor, Lora Aroyo, Zhufeng Pan, Zachary Nado, Jakub Sygnowski, Stephanie Winkler, Dian Yu, Mohammad Saleh, Loren Maggiore, Yamini Bansal, Xavier Garcia, Mehran Kazemi, Piyush Patil, Ishita Dasgupta, Iain Barr, Minh Giang, Thais Kagohara, Ivo Danihelka, Amit Marathe, Vladimir Feinberg, Mohamed Elhawaty, Nimesh Ghelani, Dan Horgan, Helen Miller, Lexi Walker, Richard Tanburn, Mukarram Tariq, Disha Shrivastava, Fei Xia, Qingze Wang, Chung-Cheng Chiu, Zoe Ashwood, Khuslen Baatarsukh, Sina Samangooei, Rapha\"el Lopez Kaufman, Fred Alcober, Axel Stjerngren, Paul Komarek, Katerina Tsihlas, Anudhyan Boral, Ramona Comanescu, Jeremy Chen, Ruibo Liu, Chris Welty, Dawn Bloxwich, Charlie Chen, Yanhua Sun, Fangxiaoyu Feng, Matthew Mauger, Xerxes Dotiwalla, Vincent Hellendoorn, Michael Sharman, Ivy Zheng, Krishna Haridasan, Gabe Barth-Maron, Craig Swanson, Dominika Rogozi\'nska, Alek Andreev, Paul Kishan Rubenstein, Ruoxin Sang, Dan Hurt, Gamaleldin Elsayed, Renshen Wang, Dave Lacey, Anastasija Ili\'c, Yao Zhao, Adam Iwanicki, Alejandro Lince, Alexander Chen, Christina Lyu, Carl Lebsack, Jordan Griffith, Meenu Gaba, Paramjit Sandhu, Phil Chen, Anna Koop, Ravi Rajwar, Soheil Hassas Yeganeh, Solomon Chang, Rui Zhu, Soroush Radpour, Elnaz Davoodi, Ving Ian Lei, Yang Xu, Daniel Toyama, Constant Segal, Martin Wicke, Hanzhao Lin, Anna Bulanova, Adri\`a Puigdom\`enech Badia, Nemanja Raki\'cevi\'c, Pablo Sprechmann, Angelos Filos, Shaobo Hou, V\'ictor Campos, Nora Kassner, Devendra Sachan, Meire Fortunato, Chimezie Iwuanyanwu, Vitaly Nikolaev, Balaji Lakshminarayanan, Sadegh Jazayeri, Mani Varadarajan, Chetan Tekur, Doug Fritz, Misha Khalman, David Reitter, Kingshuk Dasgupta, Shourya Sarcar, Tina Ornduff, Javier Snaider, Fantine Huot, Johnson Jia, Rupert Kemp, Nejc Trdin, Anitha Vijayakumar, Lucy Kim, Christof Angermueller, Li Lao, Tianqi Liu, Haibin Zhang, David Engel, Somer Greene, Ana\"is White, Jessica Austin, Lilly Taylor, Shereen Ashraf, Dangyi Liu, Maria Georgaki, Irene Cai, Yana Kulizhskaya, Sonam Goenka, Brennan Saeta, Ying Xu, Christian Frank, Dario de Cesare, Brona Robenek, Harry Richardson, Mahmoud Alnahlawi, Christopher Yew, Priya Ponnapalli, Marco Tagliasacchi, Alex Korchemniy, Yelin Kim, Dinghua Li, Bill Rosgen, Kyle Levin, Jeremy Wiesner, Praseem Banzal, Praveen Srinivasan, Hongkun Yu, \c{C}a\u{g}lar \"Unl\"u, David Reid, Zora Tung, Daniel Finchelstein, Ravin Kumar, Andre Elisseeff, Jin Huang, Ming Zhang, Ricardo Aguilar, Mai Gim\'enez, Jiawei Xia, Olivier Dousse, Willi Gierke, Damion Yates, Komal Jalan, Lu Li, Eri Latorre-Chimoto, Duc Dung Nguyen, Ken Durden, Praveen Kallakuri, Yaxin Liu, Matthew Johnson, Tomy Tsai, Alice Talbert, Jasmine Liu, Alexander Neitz, Chen Elkind, Marco Selvi, Mimi Jasarevic, Livio Baldini Soares, Albert Cui, Pidong Wang, Alek Wenjiao Wang, Xinyu Ye, Krystal Kallarackal, Lucia Loher, Hoi Lam, Josef Broder, Dan Holtmann-Rice, Nina Martin, Bramandia Ramadhana, Mrinal Shukla, Sujoy Basu, Abhi Mohan, Nick Fernando, Noah Fiedel, Kim Paterson, Hui Li, Ankush Garg, Jane Park, DongHyun Choi, Diane Wu, Sankalp Singh, Zhishuai Zhang, Amir Globerson, Lily Yu, John Carpenter, F\'elix de Chaumont Quitry, Carey Radebaugh, Chu-Cheng Lin, Alex Tudor, Prakash Shroff, Drew Garmon, Dayou Du, Neera Vats, Han Lu, Shariq Iqbal, Alex Yakubovich, Nilesh Tripuraneni, James Manyika, Haroon Qureshi, Nan Hua, Christel Ngani, Maria Abi Raad, Hannah Forbes, Jeff Stanway, Mukund Sundararajan, Victor Ungureanu, Colton Bishop, Yunjie Li, Balaji Venkatraman, Bo Li, Chloe Thornton, Salvatore Scellato, Nishesh Gupta, Yicheng Wang, Ian Tenney, Xihui Wu, Ashish Shenoy, Gabriel Carvajal, Diana Gage Wright, Ben Bariach, Zhuyun Xiao, Peter Hawkins, Sid Dalmia, Clement Farabet, Pedro Valenzuela, Quan Yuan, Ananth Agarwal, Mia Chen, Wooyeol Kim, Brice Hulse, Nandita Dukkipati, Adam Paszke, Andrew Bolt, Kiam Choo, Jennifer Beattie, Jennifer Prendki, Harsha Vashisht, Rebeca Santamaria-Fernandez, Luis C. Cobo, Jarek Wilkiewicz, David Madras, Ali Elqursh, Grant Uy, Kevin Ramirez, Matt Harvey, Tyler Liechty, Heiga Zen, Jeff Seibert, Clara Huiyi Hu, Andrey Khorlin, Maigo Le, Asaf Aharoni, Megan Li, Lily Wang, Sandeep Kumar, Norman Casagrande, Jay Hoover, Dalia El Badawy, David Soergel, Denis Vnukov, Matt Miecnikowski, Jiri Simsa, Praveen Kumar, Thibault Sellam, Daniel Vlasic, Samira Daruki, Nir Shabat, John Zhang, Guolong Su, Jiageng Zhang, Jeremiah Liu, Yi Sun, Evan Palmer, Alireza Ghaffarkhah, Xi Xiong, Victor Cotruta, Michael Fink, Lucas Dixon, Ashwin Sreevatsa, Adrian Goedeckemeyer, Alek Dimitriev, Mohsen Jafari, Remi Crocker, Nicholas FitzGerald, Aviral Kumar, Sanjay Ghemawat, Ivan Philips, Frederick Liu, Yannie Liang, Rachel Sterneck, Alena Repina, Marcus Wu, Laura Knight, Marin Georgiev, Hyo Lee, Harry Askham, Abhishek Chakladar, Annie Louis, Carl Crous, Hardie Cate, Dessie Petrova, Michael Quinn, Denese Owusu-Afriyie, Achintya Singhal, Nan Wei, Solomon Kim, Damien Vincent, Milad Nasr, Christopher A. Choquette-Choo, Reiko Tojo, Shawn Lu, Diego de Las Casas, Yuchung Cheng, Tolga Bolukbasi, Katherine Lee, Saaber Fatehi, Rajagopal Ananthanarayanan, Miteyan Patel, Charbel Kaed, Jing Li, Shreyas Rammohan Belle, Zhe Chen, Jaclyn Konzelmann, Siim P\~oder, Roopal Garg, Vinod Koverkathu, Adam Brown, Chris Dyer, Rosanne Liu, Azade Nova, Jun Xu, Alanna Walton, Alicia Parrish, Mark Epstein, Sara McCarthy, Slav Petrov, Demis Hassabis, Koray Kavukcuoglu, Jeffrey Dean, Oriol Vinyals
Abstract: In this report, we present the latest model of the Gemini family, Gemini 1.5 Pro, a highly compute-efficient multimodal mixture-of-experts model capable of recalling and reasoning over fine-grained information from millions of tokens of context, including multiple long documents and hours of video and audio. Gemini 1.5 Pro achieves near-perfect recall on long-context retrieval tasks across modalities, improves the state-of-the-art in long-document QA, long-video QA and long-context ASR, and matches or surpasses Gemini 1.0 Ultra's state-of-the-art performance across a broad set of benchmarks. Studying the limits of Gemini 1.5 Pro's long-context ability, we find continued improvement in next-token prediction and near-perfect retrieval (>99%) up to at least 10M tokens, a generational leap over existing models such as Claude 2.1 (200k) and GPT-4 Turbo (128k). Finally, we highlight surprising new capabilities of large language models at the frontier; when given a grammar manual for Kalamang, a language with fewer than 200 speakers worldwide, the model learns to translate English to Kalamang at a similar level to a person who learned from the same content.
Authors: Mengsha Liu, Daoyuan Chen, Yaliang Li, Guian Fang, Ying Shen
Abstract: Data visualization serves as a critical means for presenting data and mining its valuable insights. The task of chart summarization, through natural language processing techniques, facilitates in-depth data analysis of charts. However, there still are notable deficiencies in terms of visual-language matching and reasoning ability for existing approaches. To address these limitations, this study constructs a large-scale dataset of comprehensive chart-caption pairs and fine-tuning instructions on each chart. Thanks to the broad coverage of various topics and visual styles within this dataset, better matching degree can be achieved from the view of training data. Moreover, we propose an innovative chart summarization method, ChartThinker, which synthesizes deep analysis based on chains of thought and strategies of context retrieval, aiming to improve the logical coherence and accuracy of the generated summaries. Built upon the curated datasets, our trained model consistently exhibits superior performance in chart summarization tasks, surpassing 8 state-of-the-art models over 7 evaluation metrics. Our dataset and codes are publicly accessible.
Authors: Dawei Zhu, Liang Wang, Nan Yang, Yifan Song, Wenhao Wu, Furu Wei, Sujian Li
Abstract: Embedding models play a pivot role in modern NLP applications such as IR and RAG. While the context limit of LLMs has been pushed beyond 1 million tokens, embedding models are still confined to a narrow context window not exceeding 8k tokens, refrained from application scenarios requiring long inputs such as legal contracts. This paper explores context window extension of existing embedding models, pushing the limit to 32k without requiring additional training. First, we examine the performance of current embedding models for long context retrieval on our newly constructed LongEmbed benchmark. LongEmbed comprises two synthetic tasks and four carefully chosen real-world tasks, featuring documents of varying length and dispersed target information. Benchmarking results underscore huge room for improvement in these models. Based on this, comprehensive experiments show that training-free context window extension strategies like position interpolation can effectively extend the context window of existing embedding models by several folds, regardless of their original context being 512 or beyond 4k. Furthermore, for models employing absolute position encoding (APE), we show the possibility of further fine-tuning to harvest notable performance gains while strictly preserving original behavior for short inputs. For models using rotary position embedding (RoPE), significant enhancements are observed when employing RoPE-specific methods, such as NTK and SelfExtend, indicating RoPE's superiority over APE for context window extension. To facilitate future research, we release E5-Base-4k and E5-RoPE-Base, along with the LongEmbed benchmark.
Authors: Yanzeng Li, Cheng Zeng, Jialun Zhong, Ruoyu Zhang, Minhao Zhang, Lei Zou
Abstract: Simulated Patients (SPs) play a crucial role in clinical medical education by providing realistic scenarios for student practice. However, the high cost of training and hiring qualified SPs, along with the heavy workload and potential risks they face in consistently portraying actual patients, limit students' access to this type of clinical training. Consequently, the integration of computer program-based simulated patients has emerged as a valuable educational tool in recent years. With the rapid development of Large Language Models (LLMs), their exceptional capabilities in conversational artificial intelligence and role-playing have been demonstrated, making them a feasible option for implementing Virtual Simulated Patient (VSP). In this paper, we present an integrated model-agnostic framework called CureFun that harnesses the potential of LLMs in clinical medical education. This framework facilitates natural conversations between students and simulated patients, evaluates their dialogue, and provides suggestions to enhance students' clinical inquiry skills. Through comprehensive evaluations, our approach demonstrates more authentic and professional SP-scenario dialogue flows compared to other LLM-based chatbots, thus proving its proficiency in simulating patients. Additionally, leveraging CureFun's evaluation ability, we assess several medical LLMs and discuss the possibilities and limitations of using LLMs as virtual doctors from the perspective of their diagnostic abilities.
Authors: Mengzhao Jia, Zhihan Zhang, Wenhao Yu, Fangkai Jiao, Meng Jiang
Abstract: Open-source multimodal large language models (MLLMs) excel in various tasks involving textual and visual inputs but still struggle with complex multimodal mathematical reasoning, lagging behind proprietary models like GPT-4V(ision) and Gemini-Pro. Although fine-tuning with intermediate steps (i.e., rationales) elicits some mathematical reasoning skills, the resulting models still fall short in visual comprehension due to inadequate visual-centric supervision, which leads to inaccurate interpretation of math figures. To address this issue, we propose a two-step training pipeline VCAR, which emphasizes the Visual Comprehension training in Addition to mathematical Reasoning learning. It first improves the visual comprehension ability of MLLMs through the visual description generation task, followed by another training step on generating rationales with the assistance of descriptions. Experimental results on two popular benchmarks demonstrate that VCAR substantially outperforms baseline methods solely relying on rationale supervision, especially on problems with high visual demands.
Authors: Aleksi Huotala, Miikka Kuutila, Paul Ralph, Mika M\"antyl\"a
Abstract: Systematic review (SR) is a popular research method in software engineering (SE). However, conducting an SR takes an average of 67 weeks. Thus, automating any step of the SR process could reduce the effort associated with SRs. Our objective is to investigate if Large Language Models (LLMs) can accelerate title-abstract screening by simplifying abstracts for human screeners, and automating title-abstract screening. We performed an experiment where humans screened titles and abstracts for 20 papers with both original and simplified abstracts from a prior SR. The experiment with human screeners was reproduced with GPT-3.5 and GPT-4 LLMs to perform the same screening tasks. We also studied if different prompting techniques (Zero-shot (ZS), One-shot (OS), Few-shot (FS), and Few-shot with Chain-of-Thought (FS-CoT)) improve the screening performance of LLMs. Lastly, we studied if redesigning the prompt used in the LLM reproduction of screening leads to improved performance. Text simplification did not increase the screeners' screening performance, but reduced the time used in screening. Screeners' scientific literacy skills and researcher status predict screening performance. Some LLM and prompt combinations perform as well as human screeners in the screening tasks. Our results indicate that the GPT-4 LLM is better than its predecessor, GPT-3.5. Additionally, Few-shot and One-shot prompting outperforms Zero-shot prompting. Using LLMs for text simplification in the screening process does not significantly improve human performance. Using LLMs to automate title-abstract screening seems promising, but current LLMs are not significantly more accurate than human screeners. To recommend the use of LLMs in the screening process of SRs, more research is needed. We recommend future SR studies publish replication packages with screening data to enable more conclusive experimenting with LLM screening.
Authors: Yanqi Zhou, Nan Du, Yanping Huang, Daiyi Peng, Chang Lan, Da Huang, Siamak Shakeri, David So, Andrew Dai, Yifeng Lu, Zhifeng Chen, Quoc Le, Claire Cui, James Laudon, Jeff Dean
Abstract: Transformers are central to recent successes in natural language processing and computer vision. Transformers have a mostly uniform backbone where layers alternate between feed-forward and self-attention in order to build a deep network. Here we investigate this design choice and find that more complex blocks that have different permutations of layer primitives can be more efficient. Using this insight, we develop a complex block, named Brainformer, that consists of a diverse sets of layers such as sparsely gated feed-forward layers, dense feed-forward layers, attention layers, and various forms of layer normalization and activation functions. Brainformer consistently outperforms the state-of-the-art dense and sparse Transformers, in terms of both quality and efficiency. A Brainformer model with 8 billion activated parameters per token demonstrates 2x faster training convergence and 5x faster step time compared to its GLaM counterpart. In downstream task evaluation, Brainformer also demonstrates a 3% higher SuperGLUE score with fine-tuning compared to GLaM with a similar number of activated parameters. Finally, Brainformer largely outperforms a Primer dense model derived with NAS with similar computation per token on fewshot evaluations.
Authors: Zhaocheng Zhu, Yuan Xue, Xinyun Chen, Denny Zhou, Jian Tang, Dale Schuurmans, Hanjun Dai
Abstract: When prompted with a few examples and intermediate steps, large language models (LLMs) have demonstrated impressive performance in various reasoning tasks. However, prompting methods that rely on implicit knowledge in an LLM often generate incorrect answers when the implicit knowledge is wrong or inconsistent with the task. To tackle this problem, we present Hypotheses-to-Theories (HtT), a framework that learns a rule library for reasoning with LLMs. HtT contains two stages, an induction stage and a deduction stage. In the induction stage, an LLM is first asked to generate and verify rules over a set of training examples. Rules that appear and lead to correct answers sufficiently often are collected to form a rule library. In the deduction stage, the LLM is then prompted to employ the learned rule library to perform reasoning to answer test questions. Experiments on relational reasoning, numerical reasoning and concept learning problems show that HtT improves existing prompting methods, with an absolute gain of 10-30% in accuracy. The learned rules are also transferable to different models and to different forms of the same problem.
Authors: Chang Liu, Yuanhe Tian, Yan Song
Abstract: Radiology report generation (RRG) aims to automatically generate free-text descriptions from clinical radiographs, e.g., chest X-Ray images. RRG plays an essential role in promoting clinical automation and presents significant help to provide practical assistance for inexperienced doctors and alleviate radiologists' workloads. Therefore, consider these meaningful potentials, research on RRG is experiencing explosive growth in the past half-decade, especially with the rapid development of deep learning approaches. Existing studies perform RRG from the perspective of enhancing different modalities, provide insights on optimizing the report generation process with elaborated features from both visual and textual information, and further facilitate RRG with the cross-modal interactions among them. In this paper, we present a comprehensive review of deep learning-based RRG from various perspectives. Specifically, we firstly cover pivotal RRG approaches based on the task-specific features of radiographs, reports, and the cross-modal relations between them, and then illustrate the benchmark datasets conventionally used for this task with evaluation metrics, subsequently analyze the performance of different approaches and finally offer our summary on the challenges and the trends in future directions. Overall, the goal of this paper is to serve as a tool for understanding existing literature and inspiring potential valuable research in the field of RRG.
Authors: Kangwook Jang, Sungnyun Kim, Hoirin Kim
Abstract: Albeit great performance of Transformer-based speech selfsupervised learning (SSL) models, their large parameter size and computational cost make them unfavorable to utilize. In this study, we propose to compress the speech SSL models by distilling speech temporal relation (STaR). Unlike previous works that directly match the representation for each speech frame, STaR distillation transfers temporal relation between speech frames, which is more suitable for lightweight student with limited capacity. We explore three STaR distillation objectives and select the best combination as the final STaR loss. Our model distilled from HuBERT BASE achieves an overall score of 79.8 on SUPERB benchmark, the best performance among models with up to 27 million parameters. We show that our method is applicable across different speech SSL models and maintains robust performance with further reduced parameters.
Authors: Di Zhang, Wei Liu, Qian Tan, Jingdan Chen, Hang Yan, Yuliang Yan, Jiatong Li, Weiran Huang, Xiangyu Yue, Wanli Ouyang, Dongzhan Zhou, Shufei Zhang, Mao Su, Han-Sen Zhong, Yuqiang Li
Abstract: Large language models (LLMs) have made impressive progress in chemistry applications. However, the community lacks an LLM specifically designed for chemistry. The main challenges are two-fold: firstly, most chemical data and scientific knowledge are stored in structured databases, which limits the model's ability to sustain coherent dialogue when used directly. Secondly, there is an absence of objective and fair benchmark that encompass most chemistry tasks. Here, we introduce ChemLLM, a comprehensive framework that features the first LLM dedicated to chemistry. It also includes ChemData, a dataset specifically designed for instruction tuning, and ChemBench, a robust benchmark covering nine essential chemistry tasks. ChemLLM is adept at performing various tasks across chemical disciplines with fluid dialogue interaction. Notably, ChemLLM achieves results comparable to GPT-4 on the core chemical tasks and demonstrates competitive performance with LLMs of similar size in general scenarios. ChemLLM paves a new path for exploration in chemical studies, and our method of incorporating structured chemical knowledge into dialogue systems sets a new standard for developing LLMs in various scientific fields. Codes, Datasets, and Model weights are publicly accessible at https://hf.co/AI4Chem
URLs: https://hf.co/AI4Chem
Authors: Nathaniel Li, Alexander Pan, Anjali Gopal, Summer Yue, Daniel Berrios, Alice Gatti, Justin D. Li, Ann-Kathrin Dombrowski, Shashwat Goel, Long Phan, Gabriel Mukobi, Nathan Helm-Burger, Rassin Lababidi, Lennart Justen, Andrew B. Liu, Michael Chen, Isabelle Barrass, Oliver Zhang, Xiaoyuan Zhu, Rishub Tamirisa, Bhrugu Bharathi, Adam Khoja, Zhenqi Zhao, Ariel Herbert-Voss, Cort B. Breuer, Samuel Marks, Oam Patel, Andy Zou, Mantas Mazeika, Zifan Wang, Palash Oswal, Weiran Liu, Adam A. Hunt, Justin Tienken-Harder, Kevin Y. Shih, Kemper Talley, John Guan, Russell Kaplan, Ian Steneker, David Campbell, Brad Jokubaitis, Alex Levinson, Jean Wang, William Qian, Kallol Krishna Karmakar, Steven Basart, Stephen Fitz, Mindy Levine, Ponnurangam Kumaraguru, Uday Tupakula, Vijay Varadharajan, Yan Shoshitaishvili, Jimmy Ba, Kevin M. Esvelt, Alexandr Wang, Dan Hendrycks
Abstract: The White House Executive Order on Artificial Intelligence highlights the risks of large language models (LLMs) empowering malicious actors in developing biological, cyber, and chemical weapons. To measure these risks of malicious use, government institutions and major AI labs are developing evaluations for hazardous capabilities in LLMs. However, current evaluations are private, preventing further research into mitigating risk. Furthermore, they focus on only a few, highly specific pathways for malicious use. To fill these gaps, we publicly release the Weapons of Mass Destruction Proxy (WMDP) benchmark, a dataset of 4,157 multiple-choice questions that serve as a proxy measurement of hazardous knowledge in biosecurity, cybersecurity, and chemical security. WMDP was developed by a consortium of academics and technical consultants, and was stringently filtered to eliminate sensitive information prior to public release. WMDP serves two roles: first, as an evaluation for hazardous knowledge in LLMs, and second, as a benchmark for unlearning methods to remove such hazardous knowledge. To guide progress on unlearning, we develop CUT, a state-of-the-art unlearning method based on controlling model representations. CUT reduces model performance on WMDP while maintaining general capabilities in areas such as biology and computer science, suggesting that unlearning may be a concrete path towards reducing malicious use from LLMs. We release our benchmark and code publicly at https://wmdp.ai
URLs: https://wmdp.ai
Authors: Jen-tse Huang, Eric John Li, Man Ho Lam, Tian Liang, Wenxuan Wang, Youliang Yuan, Wenxiang Jiao, Xing Wang, Zhaopeng Tu, Michael R. Lyu
Abstract: Decision-making, a complicated task requiring various types of abilities, presents an excellent framework for assessing Large Language Models (LLMs). Our research investigates LLMs' decision-making capabilities through the lens of a well-established field, Game Theory. We focus specifically on games that support the participation of more than two agents simultaneously. Subsequently, we introduce our framework, GAMA-Bench, including eight classical multi-agent games. We design a scoring scheme to assess a model's performance in these games quantitatively. Through GAMA-Bench, we investigate LLMs' robustness, generalizability, and enhancement strategies. Results reveal that while GPT-3.5 shows satisfying robustness, its generalizability is relatively limited. However, its performance can be improved through approaches such as Chain-of-Thought. Additionally, we conduct evaluations across various LLMs and find that GPT-4 outperforms other models on GAMA-Bench, achieving a score of 60.5. Moreover, Gemini-1.0-Pro and GPT-3.5 (0613, 1106, 0125) demonstrate similar intelligence on GAMA-Bench. The code and experimental results are made publicly available via https://github.com/CUHK-ARISE/GAMABench.
Authors: Xingyu Fu, Yushi Hu, Bangzheng Li, Yu Feng, Haoyu Wang, Xudong Lin, Dan Roth, Noah A. Smith, Wei-Chiu Ma, Ranjay Krishna
Abstract: We introduce Blink, a new benchmark for multimodal language models (LLMs) that focuses on core visual perception abilities not found in other evaluations. Most of the Blink tasks can be solved by humans "within a blink" (e.g., relative depth estimation, visual correspondence, forensics detection, and multi-view reasoning). However, we find these perception-demanding tasks cast significant challenges for current multimodal LLMs because they resist mediation through natural language. Blink reformats 14 classic computer vision tasks into 3,807 multiple-choice questions, paired with single or multiple images and visual prompting. While humans get 95.70% accuracy on average, Blink is surprisingly challenging for existing multimodal LLMs: even the best-performing GPT-4V and Gemini achieve accuracies of 51.26% and 45.72%, only 13.17% and 7.63% higher than random guessing, indicating that such perception abilities have not "emerged" yet in recent multimodal LLMs. Our analysis also highlights that specialist CV models could solve these problems much better, suggesting potential pathways for future improvements. We believe Blink will stimulate the community to help multimodal LLMs catch up with human-level visual perception.
Authors: Chao Jin, Zili Zhang, Xuanlin Jiang, Fangyue Liu, Xin Liu, Xuanzhe Liu, Xin Jin
Abstract: Retrieval-Augmented Generation (RAG) has shown significant improvements in various natural language processing tasks by integrating the strengths of large language models (LLMs) and external knowledge databases. However, RAG introduces long sequence generation and leads to high computation and memory costs. We propose RAGCache, a novel multilevel dynamic caching system tailored for RAG. Our analysis benchmarks current RAG systems, pinpointing the performance bottleneck (i.e., long sequence due to knowledge injection) and optimization opportunities (i.e., caching knowledge's intermediate states). Based on these insights, we design RAGCache, which organizes the intermediate states of retrieved knowledge in a knowledge tree and caches them in the GPU and host memory hierarchy. RAGCache proposes a replacement policy that is aware of LLM inference characteristics and RAG retrieval patterns. It also dynamically overlaps the retrieval and inference steps to minimize the end-to-end latency. We implement RAGCache and evaluate it on vLLM, a state-of-the-art LLM inference system and Faiss, a state-of-the-art vector database. The experimental results show that RAGCache reduces the time to first token (TTFT) by up to 4x and improves the throughput by up to 2.1x compared to vLLM integrated with Faiss.
Authors: Zhen Ye, Zeqian Ju, Haohe Liu, Xu Tan, Jianyi Chen, Yiwen Lu, Peiwen Sun, Jiahao Pan, Weizhen Bian, Shulin He, Qifeng Liu, Yike Guo, Wei Xue
Abstract: Recent progress in large-scale zero-shot speech synthesis has been significantly advanced by language models and diffusion models. However, the generation process of both methods is slow and computationally intensive. Efficient speech synthesis using a lower computing budget to achieve quality on par with previous work remains a significant challenge. In this paper, we present FlashSpeech, a large-scale zero-shot speech synthesis system with approximately 5\% of the inference time compared with previous work. FlashSpeech is built on the latent consistency model and applies a novel adversarial consistency training approach that can train from scratch without the need for a pre-trained diffusion model as the teacher. Furthermore, a new prosody generator module enhances the diversity of prosody, making the rhythm of the speech sound more natural. The generation processes of FlashSpeech can be achieved efficiently with one or two sampling steps while maintaining high audio quality and high similarity to the audio prompt for zero-shot speech generation. Our experimental results demonstrate the superior performance of FlashSpeech. Notably, FlashSpeech can be about 20 times faster than other zero-shot speech synthesis systems while maintaining comparable performance in terms of voice quality and similarity. Furthermore, FlashSpeech demonstrates its versatility by efficiently performing tasks like voice conversion, speech editing, and diverse speech sampling. Audio samples can be found in https://flashspeech.github.io/.
Authors: Tianshu Wang, Hongyu Lin, Xianpei Han, Xiaoyang Chen, Boxi Cao, Le Sun
Abstract: Blocking is a critical step in entity resolution, and the emergence of neural network-based representation models has led to the development of dense blocking as a promising approach for exploring deep semantics in blocking. However, previous advanced self-supervised dense blocking approaches require domain-specific training on the target domain, which limits the benefits and rapid adaptation of these methods. To address this issue, we propose UniBlocker, a dense blocker that is pre-trained on a domain-independent, easily-obtainable tabular corpus using self-supervised contrastive learning. By conducting domain-independent pre-training, UniBlocker can be adapted to various downstream blocking scenarios without requiring domain-specific fine-tuning. To evaluate the universality of our entity blocker, we also construct a new benchmark covering a wide range of blocking tasks from multiple domains and scenarios. Our experiments show that the proposed UniBlocker, without any domain-specific learning, significantly outperforms previous self- and unsupervised dense blocking methods and is comparable and complementary to the state-of-the-art sparse blocking methods.