EMNLP 2024 论文列表

Detecting hallucinations, where Large Language Models (LLMs) are not factually consistent with a Knowledge Base (KB), is a challenge for Retrieval-Augmented Generation (RAG) systems. Current solutions rely on public datasets to develop prompts or fine-tune a Natural Language Inference (NLI) model. However, these approaches are not focused on developing an enterprise RAG system; they do not consider latency, train or evaluate on production data, nor do they handle non-verifiable statements such as small talk or questions. To address this, we leverage the customer service conversation data of four large brands to evaluate existing solutions and propose a set of small encoder models trained on a new dataset. We find the proposed models to outperform existing methods and highlight the value of combining a small amount of in-domain data with public datasets.

The concept of *persona*, originally adopted in dialogue literature, has re-surged as a promising framework for tailoring large language models (LLMs) to specific context (*e.g.*, personalized search, LLM-as-a-judge). However, the growing research on leveraging persona in LLMs is relatively disorganized and lacks a systematic taxonomy. To close the gap, we present a comprehensive survey to categorize the current state of the field. We identify two lines of research, namely (1) *LLM Role-Playing*, where personas are assigned to LLMs, and (2) *LLM Personalization*, where LLMs take care of user personas. Additionally, we introduce existing methods for LLM personality evaluation. To the best of our knowledge, we present the first survey for role-playing and personalization in LLMs under the unified view of persona. We continuously maintain a paper collection to foster future endeavors.

Current automatic speech recognition (ASR) models are designed to be used across many languages and tasks without substantial changes. However, this broad language coverage hides performance gaps within languages, for example, across genders. Our study systematically evaluates the performance of two widely used multilingual ASR models on three datasets, encompassing 19 languages from eight language families and two speaking conditions. Our findings reveal clear gender disparities, with the advantaged group varying across languages and models. Surprisingly, those gaps are not explained by acoustic or lexical properties. However, probing internal model states reveals a correlation with gendered performance gap. That is, the easier it is to distinguish speaker gender in a language using probes, the more the gap reduces, favoring female speakers. Our results show that gender disparities persist even in state-of-the-art models. Our findings have implications for the improvement of multilingual ASR systems, underscoring the importance of accessibility to training data and nuanced evaluation to predict and mitigate gender gaps. We release all code and artifacts at https://github.com/g8a9/multilingual-asr-gender-gap.

There has been a growing body of work focusing on the in-context learning (ICL) abilities of large language models (LLMs). However, it is an open question how effective ICL can be. This paper presents Tutor-ICL, a simple prompting method for classification tasks inspired by how effective instructors might engage their students in learning a task. Specifically, we propose presenting exemplar answers in a *comparative format* rather than the traditional single-answer format. We also show that including the test instance before the exemplars can improve performance, making it easier for LLMs to focus on relevant exemplars. Lastly, we include a summarization step before attempting the test, following a common human practice. Experiments on various classification tasks, conducted across both decoder-only LLMs (Llama 2, 3) and encoder-decoder LLMs (Flan-T5-XL, XXL), show that Tutor-ICL consistently boosts performance, achieving up to a 13.76% increase in accuracy.

The vast majority of today’s large language models (LLMs) are English-centric, having been pretrained predominantly on English text. Yet, in order to meet user expectations, models need to be able to respond appropriately in multiple languages once deployed in downstream applications. This requires strong cross-lingual transfer abilities.In this work, we investigate the minimal amount of multilinguality required during finetuning to elicit cross-lingual generalisation in English-centric LLMs. In experiments across four LLMs, we find that multilingual instruction tuning with as few as two to three languages is both necessary and sufficient to elicit effective cross-lingual generalisation, with the limiting factor being the degree to which a target language is seen during pretraining. Evaluations on five different tasks further reveal that multilingual instruction tuning is most beneficial for generative tasks that assume input/output language agreement, such as in chat settings, while being of less importance for highly structured classification-style tasks.

Sparse Mixture of Experts (MoE) models are popular for training large language models due to their computational efficiency. However, the commonly used top-k routing mechanism suffers from redundancy computation and memory costs due to the unbalanced routing. Some experts are overflow, where the exceeding tokens are dropped. While some experts are empty, which are padded with zeros, negatively impacting model performance. To address the dropped tokens and padding, we propose the Rectify-Router, comprising the Intra-GPU Rectification and the Fill-in Rectification. The Intra-GPU Rectification handles dropped tokens, efficiently routing them to experts within the GPU where they are located to avoid inter-GPU communication. The Fill-in Rectification addresses padding by replacing padding tokens with the tokens that have high routing scores. Our experimental results demonstrate that the Intra-GPU Rectification and the Fill-in Rectification effectively handle dropped tokens and padding, respectively. Furthermore, the combination of them achieves superior performance, surpassing the accuracy of the vanilla top-1 router by 4.7%.

Multiple choice question answering tasks evaluate the reasoning, comprehension, and mathematical abilities of Large Language Models (LLMs). While existing benchmarks employ automatic translation for multilingual evaluation, this approach is error-prone and potentially introduces culturally biased questions, especially in social sciences. We introduce the first multitask, multiple-choice Turkish QA benchmark, TurkishMMLU, to evaluate LLMs’ understanding of the Turkish language. TurkishMMLU includes over 10,000 questions, covering 9 different subjects from Turkish high-school education curricula. These questions are written by curriculum experts, suitable for the high-school curricula in Turkey, covering subjects ranging from natural sciences and math questions to more culturally representative topics such as Turkish Literature and the history of the Turkish Republic. We evaluate over 20 LLMs, including multilingual open-source (e.g., Gemma, Llama, MT5), closed-source (GPT 4o, Claude, Gemini), and Turkish-adapted (e.g., Trendyol) models. We provide an extensive evaluation, including zero-shot and few-shot evaluation of LLMs, chain-of-thought reasoning, and question difficulty analysis along with model performance. We provide an in-depth analysis of the Turkish capabilities and limitations of current LLMs to provide insights for future LLMs for the Turkish language. We publicly release our code for the dataset and evaluation: https://github.com/ArdaYueksel/TurkishMMLU

We present TuringQ, the first benchmark designed to evaluate the reasoning capabilities of large language models (LLMs) in the theory of computation. TuringQ consists of 4,006 undergraduate and graduate-level question-answer pairs, categorized into four difficulty levels and covering seven core theoretical areas. We evaluate several open-source LLMs, as well as GPT-4, using Chain of Thought prompting and expert human assessment. Additionally, we propose an automated LLM-based evaluation system that demonstrates competitive accuracy when compared to human evaluation. Fine-tuning a Llama3-8B model on TuringQ shows measurable improvements in reasoning ability and out-of-domain tasks such as algebra. TuringQ serves as both a benchmark and a resource for enhancing LLM performance in complex computational reasoning tasks. Our analysis offers insights into LLM capabilities and advances in AI comprehension of theoretical computer science.

Document assistant chatbots are empowered with extensive capabilities by Large Language Models (LLMs) and have exhibited significant advancements. However, these systems may suffer from hallucinations that are difficult to verify in the context of given documents.Moreover, despite the emergence of products for document assistants, they either heavily rely on commercial LLM APIs or lack transparency in their technical implementations, leading to expensive usage costs and data privacy concerns. In this work, we introduce a fully open-source document assistant chatbot with reliable attribution, named TruthReader, utilizing adapted conversational retriever and LLMs. Our system enables the LLMs to generate answers with detailed inline citations, which can be attributed to the original document paragraphs, facilitating the verification of the factual consistency of the generated text. To further adapt the generative model, we develop a comprehensive pipeline consisting of data construction and model optimization processes.This pipeline equips the LLMs with the necessary capabilities to generate accurate answers, produce reliable citations, and refuse unanswerable questions. Our codebase, data and models are released, and the video demonstration of our system is available at https://youtu.be/RYVt3itzUQM.

The rise of LLM-based agents shows great potential to revolutionize task planning, capturing significant attention. Given that these agents will be integrated into high-stake domains, ensuring their reliability and safety is crucial. This paper presents an Agent-Constitution-based agent framework, TrustAgent, with a particular focus on improving the LLM-based agent safety. The proposed framework ensures strict adherence to the Agent Constitution through three strategic components: pre-planning strategy which injects safety knowledge to the model before plan generation, in-planning strategy which enhances safety during plan generation, and post-planning strategy which ensures safety by post-planning inspection. Our experimental results demonstrate that the proposed framework can effectively enhance an LLM agent’s safety across multiple domains by identifying and mitigating potential dangers during the planning. Further analysis reveals that the framework not only improves safety but also enhances the helpfulness of the agent. Additionally, we highlight the importance of the LLM reasoning ability in adhering to the Constitution. This paper sheds light on how to ensure the safe integration of LLM-based agents into human-centric environments. Data and code are available at https://anonymous.4open.science/r/TrustAgent-06DC.

Large language and vision models (LLVMs) have been driven by the generalization power of large language models (LLMs) and the advent of visual instruction tuning. Along with scaling them up directly, these models enable LLVMs to showcase powerful vision language (VL) performances by covering diverse tasks via natural language instructions. However, existing open-source LLVMs that perform comparably to closed-source LLVMs such as GPT-4V are often considered too large (e.g., 26B, 34B, and 110B parameters), having a larger number of layers. These large models demand costly, high-end resources for both training and inference. To address this issue, we present a new efficient LLVM family with 1.8B, 3.8B, and 7B LLM model sizes, Traversal of Layers (TroL), which enables the reuse of layers in a token-wise manner. This layer traversing technique simulates the effect of looking back and retracing the answering stream while increasing the number of forward propagation layers without physically adding more layers. We demonstrate that TroL employs a simple layer traversing approach yet efficiently outperforms the open-source LLVMs with larger model sizes and rivals the performances of the closed-source LLVMs with substantial sizes.

The global escalation in emergency department patient visits poses significant challenges to efficient clinical management, particularly in clinical triage. Traditionally managed by human professionals, clinical triage is susceptible to substantial variability and high workloads. Although large language models (LLMs) demonstrate promising reasoning and understanding capabilities, directly applying them to clinical triage remains challenging due to the complex and dynamic nature of the clinical triage task. To address these issues, we introduce TriageAgent, a novel heterogeneous multi-agent framework designed to enhance collaborative decision-making in clinical triage. TriageAgent leverages LLMs for role-playing, incorporating self-confidence and early-stopping mechanisms in multi-round discussions to improve document reasoning and classification precision for triage tasks. In addition, TriageAgent employs the medical Emergency Severity Index (ESI) handbook through a retrieval-augmented generation (RAG) approach to provide precise clinical knowledge and integrates both coarse- and fine-grained ESI-level predictions in the decision-making process. Extensive experiments demonstrate that TriageAgent outperforms state-of-the-art LLM-based methods on three clinical triage test sets. Furthermore, we have released the first public benchmark dataset for clinical triage with corresponding ESI levels and human expert performance for comparison.

Recent progress with LLM-based agents has shown promising results across various tasks. However, their use in answering questions from knowledge bases remains largely unexplored. Implementing a KBQA system using traditional methods is challenging due to the shortage of task-specific training data and the complexity of creating task-focused model structures. In this paper, we present Triad, a unified framework that utilizes an LLM-based agent with multiple roles for KBQA tasks. The agent is assigned three roles to tackle different KBQA subtasks: agent as a generalist for mastering various subtasks, as a decision maker for the selection of candidates, and as an advisor for answering questions with knowledge. Our KBQA framework is executed in four phases, involving the collaboration of the agent’s multiple roles. We evaluated the performance of our framework using three benchmark datasets, and the results show that our framework outperforms state-of-the-art systems on the LC-QuAD and YAGO-QA benchmarks, yielding F1 scores of 11.8% and 20.7%, respectively.

Large Language Models (LLMs) have demonstrated remarkable performance across multipletasks through in-context learning. For complex reasoning tasks that require step-by-step thinking, Chain-of-Thought (CoT) prompting has given impressive results, especially when combined with self-consistency. Nonetheless, some tasks remain particularly difficult for LLMs to solve. Tree of Thoughts (ToT) and Graph of Thoughts (GoT) emerged as alternatives, dividing the complex problem into paths of subproblems. In this paper, we propose Tree of Problems (ToP), a simpler version of ToT, which we hypothesise can work better for complex tasks that can be divided into identical subtasks. Our empirical results show that our approach outperforms ToT and GoT, and in addition per forms better than CoT on complex reasoning tasks. All code for this paper will be made available.

Recently, image-based Large Multimodal Models (LMMs) have made significant progress in video question-answering (VideoQA) using a frame-wise approach by leveraging large-scale pretraining in a zero-shot manner. Nevertheless, these models need to be capable of finding relevant information, extracting it, and answering the question simultaneously. Currently, existing methods perform all of these steps in a single pass without being able to adapt if insufficient or incorrect information is collected. To overcome this, we introduce a modular multi-LMM agent framework based on several agents with different roles, instructed by a Planner agent that updates its instructions using shared feedback from the other agents. Specifically, we propose TraveLER, a method that can create a plan to "**Trave**rse” through the video, ask questions about individual frames to "**L**ocate” and store key information, and then "**E**valuate” if there is enough information to answer the question. Finally, if there is not enough information, our method is able to "**R**eplan” based on its collected knowledge. Through extensive experiments, we find that the proposed TraveLER approach improves performance on several VideoQA benchmarks without the need to fine-tune on specific datasets. Our code is available at https://github.com/traveler-framework/TraveLER.

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.

With the rapid advancement of machine translation research, evaluation toolkits have become essential for benchmarking system progress. Tools like COMET and SacreBLEU offer single quality score assessments that are effective for pairwise system comparisons. However, these tools provide limited insights for fine-grained system-level comparisons and the analysis of instance-level defects. To address these limitations, we introduce Translation Canvas, an explainable interface designed to pinpoint and analyze translation systems’ performance: 1) Translation Canvas assists machine translation researchers in comprehending system-level model performance by identifying common errors (their frequency and severity) and analyzing relationships between different systems based on various evaluation metrics. 2) It supports fine-grained analysis by highlighting error spans with explanations and selectively displaying systems’ predictions. According to human evaluation, Translation Canvas demonstrates superior performance over COMET and SacreBLEU packages under enjoybility and understandbility criteria.

Recent models in cross-lingual semantic role labeling (SRL) barely analyze the applicability of their network selection.We believe that network selection is important since it affects the transferability of cross-lingual models, i.e., how the model can extract universal features from source languages to label target languages.Therefore, we comprehensively compare the transferability of different graph neural network (GNN)-based models enriched with universal dependency trees.GNN-based models include transformer-based, graph convolutional network-based, and graph attention network (GAT)-based models.We focus our study on a zero-shot setting by training the models in English and evaluating the models in 23 target languages provided by the Universal Proposition Bank.Based on our experiments, we consistently show that syntax from universal dependency trees is essential for cross-lingual SRL models to achieve better transferability.Dependency-aware self-attention with relative position representations (SAN-RPRs) transfer best across languages, especially in the long-range dependency distance.We also show that dependency-aware two-attention relational GATs transfer better than SAN-RPRs in languages where most arguments lie in a 1-2 dependency distance.

Task-oriented dialogue (TOD) systems aim to efficiently handle task-oriented conversations, including information collection. How to utilize TOD accurately, efficiently and effectively for information collection has always been a critical and challenging task. Recent studies have demonstrated that Large Language Models (LLMs) excel in dialogue, instruction generation, and reasoning, and can significantly enhance the performance of TOD through fine-tuning. However, current datasets primarily cater to user-led systems and are limited to predefined specific scenarios and slots, thereby necessitating improvements in the proactiveness, diversity, and capabilities of TOD. In this study, we present a detailed multi-domain task-oriented data construction process for conversations, and a Chinese dialogue dataset generated based on this process, **TransferTOD**, which authentically simulates human-computer dialogues in 30 popular life service scenarios. Leveraging this dataset, we trained a model using full-parameter fine-tuning called **TransferTOD-7B**, showcasing notable abilities in slot filling and questioning. Our work has demonstrated its strong generalization capabilities in various downstream scenarios, significantly enhancing both data utilization efficiency and system performance. The data is released in https://github.com/KongLongGeFDU/TransferTOD.