Wenqiang Lei

Wenqiang Lei, Yao Zhang, Feifan Song et al. · pku

Proactive dialogue system is able to lead the conversation to a goal topic and has advantaged potential in bargain, persuasion and negotiation. Current corpus-based learning manner limits its practical application in real-world scenarios. To this end, we contribute to advance the study of the proactive dialogue policy to a more natural and challenging setting, i.e., interacting dynamically with users. Further, we call attention to the non-cooperative user behavior -- the user talks about off-path topics when he/she is not satisfied with the previous topics introduced by the agent. We argue that the targets of reaching the goal topic quickly and maintaining a high user satisfaction are not always converge, because the topics close to the goal and the topics user preferred may not be the same. Towards this issue, we propose a new solution named I-Pro that can learn Proactive policy in the Interactive setting. Specifically, we learn the trade-off via a learned goal weight, which consists of four factors (dialogue turn, goal completion difficulty, user satisfaction estimation, and cooperative degree). The experimental results demonstrate I-Pro significantly outperforms baselines in terms of effectiveness and interpretability.

Fengbin Zhu, Wenqiang Lei, Fuli Feng et al.

Document Visual Question Answering (VQA) aims to understand visually-rich documents to answer questions in natural language, which is an emerging research topic for both Natural Language Processing and Computer Vision. In this work, we introduce a new Document VQA dataset, named TAT-DQA, which consists of 3,067 document pages comprising semi-structured table(s) and unstructured text as well as 16,558 question-answer pairs by extending the TAT-QA dataset. These documents are sampled from real-world financial reports and contain lots of numbers, which means discrete reasoning capability is demanded to answer questions on this dataset. Based on TAT-DQA, we further develop a novel model named MHST that takes into account the information in multi-modalities, including text, layout and visual image, to intelligently address different types of questions with corresponding strategies, i.e., extraction or reasoning. Extensive experiments show that the MHST model significantly outperforms the baseline methods, demonstrating its effectiveness. However, the performance still lags far behind that of expert humans. We expect that our new TAT-DQA dataset would facilitate the research on deep understanding of visually-rich documents combining vision and language, especially for scenarios that require discrete reasoning. Also, we hope the proposed model would inspire researchers to design more advanced Document VQA models in future. Our dataset will be publicly available for non-commercial use at https://nextplusplus.github.io/TAT-DQA/.

Yang Deng, Wenxuan Zhang, Weiwen Xu et al.

Recent years witnessed several advances in developing multi-goal conversational recommender systems (MG-CRS) that can proactively attract users' interests and naturally lead user-engaged dialogues with multiple conversational goals and diverse topics. Four tasks are often involved in MG-CRS, including Goal Planning, Topic Prediction, Item Recommendation, and Response Generation. Most existing studies address only some of these tasks. To handle the whole problem of MG-CRS, modularized frameworks are adopted where each task is tackled independently without considering their interdependencies. In this work, we propose a novel Unified MultI-goal conversational recommeNDer system, namely UniMIND. In specific, we unify these four tasks with different formulations into the same sequence-to-sequence (Seq2Seq) paradigm. Prompt-based learning strategies are investigated to endow the unified model with the capability of multi-task learning. Finally, the overall learning and inference procedure consists of three stages, including multi-task learning, prompt-based tuning, and inference. Experimental results on two MG-CRS benchmarks (DuRecDial and TG-ReDial) show that UniMIND achieves state-of-the-art performance on all tasks with a unified model. Extensive analyses and discussions are provided for shedding some new perspectives for MG-CRS.

Yang Deng, Wenqiang Lei, Wenxuan Zhang et al.

To facilitate conversational question answering (CQA) over hybrid contexts in finance, we present a new dataset, named PACIFIC. Compared with existing CQA datasets, PACIFIC exhibits three key features: (i) proactivity, (ii) numerical reasoning, and (iii) hybrid context of tables and text. A new task is defined accordingly to study Proactive Conversational Question Answering (PCQA), which combines clarification question generation and CQA. In addition, we propose a novel method, namely UniPCQA, to adapt a hybrid format of input and output content in PCQA into the Seq2Seq problem, including the reformulation of the numerical reasoning process as code generation. UniPCQA performs multi-task learning over all sub-tasks in PCQA and incorporates a simple ensemble strategy to alleviate the error propagation issue in the multi-task learning by cross-validating top-$k$ sampled Seq2Seq outputs. We benchmark the PACIFIC dataset with extensive baselines and provide comprehensive evaluations on each sub-task of PCQA.

Keqin Bao, Yu Wan, Dayiheng Liu et al.

In this paper, we present our submission to the sentence-level MQM benchmark at Quality Estimation Shared Task, named UniTE (Unified Translation Evaluation). Specifically, our systems employ the framework of UniTE, which combined three types of input formats during training with a pre-trained language model. First, we apply the pseudo-labeled data examples for the continuously pre-training phase. Notably, to reduce the gap between pre-training and fine-tuning, we use data pruning and a ranking-based score normalization strategy. For the fine-tuning phase, we use both Direct Assessment (DA) and Multidimensional Quality Metrics (MQM) data from past years' WMT competitions. Finally, we collect the source-only evaluation results, and ensemble the predictions generated by two UniTE models, whose backbones are XLM-R and InfoXLM, respectively. Results show that our models reach 1st overall ranking in the Multilingual and English-Russian settings, and 2nd overall ranking in English-German and Chinese-English settings, showing relatively strong performances in this year's quality estimation competition.

Keqin Bao, Yu Wan, Dayiheng Liu et al.

Fine-grained information on translation errors is helpful for the translation evaluation community. Existing approaches can not synchronously consider error position and type, failing to integrate the error information of both. In this paper, we propose Fine-Grained Translation Error Detection (FG-TED) task, aiming at identifying both the position and the type of translation errors on given source-hypothesis sentence pairs. Besides, we build an FG-TED model to predict the \textbf{addition} and \textbf{omission} errors -- two typical translation accuracy errors. First, we use a word-level classification paradigm to form our model and use the shortcut learning reduction to relieve the influence of monolingual features. Besides, we construct synthetic datasets for model training, and relieve the disagreement of data labeling in authoritative datasets, making the experimental benchmark concordant. Experiments show that our model can identify both error type and position concurrently, and gives state-of-the-art results on the restored dataset. Our model also delivers more reliable predictions on low-resource and transfer scenarios than existing baselines. The related datasets and the source code will be released in the future.

Haofu Yang, Jiaji Liu, Chen Huang et al.

Developing non-collaborative dialogue agents traditionally requires the manual, unscalable codification of expert strategies. We propose \ours, a method that leverages large language models to autonomously induce both strategy actions and planning logic directly from raw transcripts. METRO formalizes expert knowledge into a Strategy Forest, a hierarchical structure that captures both short-term responses (nodes) and long-term strategic foresight (branches). Experimental results across two benchmarks show that METRO demonstrates promising performance, outperforming existing methods by an average of 9%-10%. Our further analysis not only reveals the success behind METRO (strategic behavioral diversity and foresight), but also demonstrates its robust cross-task transferability. This offers new insights into building non-collaborative agents in a cost-effective and scalable way. Our code is available at https://github.com/Humphrey-0125/METRO.

Mingfeng Xue, Dayiheng Liu, Kexin Yang et al. · tsinghua

The emergence of large language models (LLMs) has revolutionized natural language processing tasks. However, existing instruction-tuning datasets suffer from occupational bias: the majority of data relates to only a few occupations, which hampers the instruction-tuned LLMs to generate helpful responses to professional queries from practitioners in specific fields. To mitigate this issue and promote occupation-inclusive LLMs, we create an instruction-tuning dataset named \emph{OccuQuest}, which contains 110,000+ prompt-completion pairs and 30,000+ dialogues covering over 1,000 occupations in 26 occupational categories. We systematically request ChatGPT, organizing queries hierarchically based on Occupation, Responsibility, Topic, and Question, to ensure a comprehensive coverage of occupational specialty inquiries. By comparing with three commonly used datasets (Dolly, ShareGPT, and WizardLM), we observe that OccuQuest exhibits a more balanced distribution across occupations. Furthermore, we assemble three test sets for comprehensive evaluation, an occu-test set covering 25 occupational categories, an estate set focusing on real estate, and an occu-quora set containing real-world questions from Quora. We then fine-tune LLaMA on OccuQuest to obtain OccuLLaMA, which significantly outperforms state-of-the-art LLaMA variants (Vicuna, Tulu, and WizardLM) on professional questions in GPT-4 and human evaluations. Notably, on the occu-quora set, OccuLLaMA reaches a high win rate of 86.4\% against WizardLM.

Wentao Shi, Junkang Wu, Xuezhi Cao et al.

Learning hyperbolic embeddings for knowledge graph (KG) has gained increasing attention due to its superiority in capturing hierarchies. However, some important operations in hyperbolic space still lack good definitions, making existing methods unable to fully leverage the merits of hyperbolic space. Specifically, they suffer from two main limitations: 1) existing Graph Convolutional Network (GCN) methods in hyperbolic space rely on tangent space approximation, which would incur approximation error in representation learning, and 2) due to the lack of inner product operation definition in hyperbolic space, existing methods can only measure the plausibility of facts (links) with hyperbolic distance, which is difficult to capture complex data patterns. In this work, we contribute: 1) a Full Poincaré Multi-relational GCN that achieves graph information propagation in hyperbolic space without requiring any approximation, and 2) a hyperbolic generalization of Euclidean inner product that is beneficial to capture both hierarchical and complex patterns. On this basis, we further develop a \textbf{F}ully and \textbf{F}lexible \textbf{H}yperbolic \textbf{R}epresentation framework (\textbf{FFHR}) that is able to transfer recent Euclidean-based advances to hyperbolic space. We demonstrate it by instantiating FFHR with four representative KGC methods. Extensive experiments on benchmark datasets validate the superiority of our FFHRs over their Euclidean counterparts as well as state-of-the-art hyperbolic embedding methods.

Youcheng Huang, Wenqiang Lei, Jie Fu et al. · mila

Incorporating large-scale pre-trained models with the prototypical neural networks is a de-facto paradigm in few-shot named entity recognition. Existing methods, unfortunately, are not aware of the fact that embeddings from pre-trained models contain a prominently large amount of information regarding word frequencies, biasing prototypical neural networks against learning word entities. This discrepancy constrains the two models' synergy. Thus, we propose a one-line-code normalization method to reconcile such a mismatch with empirical and theoretical grounds. Our experiments based on nine benchmark datasets show the superiority of our method over the counterpart models and are comparable to the state-of-the-art methods. In addition to the model enhancement, our work also provides an analytical viewpoint for addressing the general problems in few-shot name entity recognition or other tasks that rely on pre-trained models or prototypical neural networks.

Weiwen Xu, Yang Deng, Wenqiang Lei et al.

We study automatic Contract Clause Extraction (CCE) by modeling implicit relations in legal contracts. Existing CCE methods mostly treat contracts as plain text, creating a substantial barrier to understanding contracts of high complexity. In this work, we first comprehensively analyze the complexity issues of contracts and distill out three implicit relations commonly found in contracts, namely, 1) Long-range Context Relation that captures the correlations of distant clauses; 2) Term-Definition Relation that captures the relation between important terms with their corresponding definitions; and 3) Similar Clause Relation that captures the similarities between clauses of the same type. Then we propose a novel framework ConReader to exploit the above three relations for better contract understanding and improving CCE. Experimental results show that ConReader makes the prediction more interpretable and achieves new state-of-the-art on two CCE tasks in both conventional and zero-shot settings.

Chen Huang, Zitan Jiang, Changyi Zou et al.

Customer service chatbots are increasingly expected to serve not merely as reactive support tools for users, but as strategic interfaces for harvesting high-value information and business intelligence. In response, we make three main contributions. 1) We introduce and define a novel task of Proactive Information Probing, which optimizes when to probe users for pre-specified target information while minimizing conversation turns and user friction. 2) We propose PROCHATIP, a proactive chatbot framework featuring a specialized conversation strategy module trained to master the delicate timing of probes. 3) Experiments demonstrate that PROCHATIP significantly outperforms baselines, exhibiting superior capability in both information probing and service quality. We believe that our work effectively redefines the commercial utility of chatbots, positioning them as scalable, cost-effective engines for proactive business intelligence. Our code is available at https://github.com/SCUNLP/PROCHATIP.

Pengfeng Li, Chen Huang, Chaoqun Hao et al.

Contextual causal reasoning is a critical yet challenging capability for Large Language Models (LLMs). Existing benchmarks, however, often evaluate this skill in fragmented settings, failing to ensure context consistency or cover the full causal hierarchy. To address this, we pioneer METER to systematically benchmark LLMs across all three levels of the causal ladder under a unified context setting. Our extensive evaluation of various LLMs reveals a significant decline in proficiency as tasks ascend the causal hierarchy. To diagnose this degradation, we conduct a deep mechanistic analysis via both error pattern identification and internal information flow tracing. Our analysis reveals two primary failure modes: (1) LLMs are susceptible to distraction by causally irrelevant but factually correct information at lower level of causality; and (2) as tasks ascend the causal hierarchy, faithfulness to the provided context degrades, leading to a reduced performance. We belive our work advances our understanding of the mechanisms behind LLM contextual causal reasoning and establishes a critical foundation for future research. Our code and dataset are available at https://github.com/SCUNLP/METER .

Kexin Yang, Dayiheng Liu, Wenqiang Lei et al.

Attribute-based Controlled Text Generation (CTG) refers to generating sentences that satisfy desirable attributes (e.g., emotions and topics). Existing works often utilize fine-tuning or resort to extra attribute classifiers, yet suffer from storage and inference time increases. To address these concerns, we explore attribute-based CTG in a prompt-based manner. In short, the proposed Tailor represents each attribute as a pre-trained continuous vector (i.e., single-attribute prompt) and guides the generation of a fixed PLM switch to a pre-specified attribute. We experimentally find that these prompts can be simply concatenated as a whole to multi-attribute CTG without any re-training, yet raises problems of fluency decrease and position sensitivity. To this end, Tailor provides a multi-attribute prompt mask and a re-indexing position-ids sequence to bridge the gap between the training (one prompt for each task) and testing stage (concatenating more than one prompt). To further enhance such single-attribute prompt combinations, Tailor also introduces a trainable prompt connector, which can be concatenated with any two single-attribute prompts to multi-attribute text generation. Experiments on 11 attribute-specific generation tasks demonstrate strong performances of Tailor on both single-attribute and multi-attribute CTG, with 0.08\% training parameters of a GPT-2.

Yu Wan, Keqin Bao, Dayiheng Liu et al.

In this report, we present our submission to the WMT 2022 Metrics Shared Task. We build our system based on the core idea of UNITE (Unified Translation Evaluation), which unifies source-only, reference-only, and source-reference-combined evaluation scenarios into one single model. Specifically, during the model pre-training phase, we first apply the pseudo-labeled data examples to continuously pre-train UNITE. Notably, to reduce the gap between pre-training and fine-tuning, we use data cropping and a ranking-based score normalization strategy. During the fine-tuning phase, we use both Direct Assessment (DA) and Multidimensional Quality Metrics (MQM) data from past years' WMT competitions. Specially, we collect the results from models with different pre-trained language model backbones, and use different ensembling strategies for involved translation directions.

Fengbin Zhu, Chao Wang, Wenqiang Lei et al.

Key Information Extraction (KIE) is aimed at extracting structured information (e.g. key-value pairs) from form-style documents (e.g. invoices), which makes an important step towards intelligent document understanding. Previous approaches generally tackle KIE by sequence tagging, which faces difficulty to process non-flatten sequences, especially for table-text mixed documents. These approaches also suffer from the trouble of pre-defining a fixed set of labels for each type of documents, as well as the label imbalance issue. In this work, we assume Optical Character Recognition (OCR) has been applied to input documents, and reformulate the KIE task as a region prediction problem in the two-dimensional (2D) space given a target field. Following this new setup, we develop a new KIE model named Region-based Document Understanding (RDU) that takes as input the text content and corresponding coordinates of a document, and tries to predict the result by localizing a bounding-box-like region. Our RDU first applies a layout-aware BERT equipped with a soft layout attention masking and bias mechanism to incorporate layout information into the representations. Then, a list of candidate regions is generated from the representations via a Region Proposal Module inspired by computer vision models widely applied for object detection. Finally, a Region Categorization Module and a Region Selection Module are adopted to judge whether a proposed region is valid and select the one with the largest probability from all proposed regions respectively. Experiments on four types of form-style documents show that our proposed method can achieve impressive results. In addition, our RDU model can be trained with different document types seamlessly, which is especially helpful over low-resource documents.

Jingwei Shi, Shengyu Tao, Xinxiang Yin et al.

The application of games as a therapeutic tool for cognitive training is beneficial for patients with cognitive impairments. However, effective game design for individual patient is resource-intensive. To this end, we propose an LLM-powered method, \ours, for automated and personalized therapeutic game design. Inspired by the Dungeons & Dragons, LETGAMES generates an open-world interactive narrative game. It not only generates game scenarios and challenges that target specific cognitive domains, but also employs conversational strategies to offer guidance and companionship. To validate its efficacy, we pioneer a psychology-grounded evaluation protocol LETGAMESEVAL, establishing comprehensive metrics for rehabilitative assessment. Building upon this, our experimental results from both LLM-based assessors and human expert evaluations demonstrate the significant potential of our approach, positioning LETGAMES as a promising solution to the widespread need for more accessible and tailored cognitive training tools. Our code will be open-sourced upon acceptance.

Peixin Qin, Chen Huang, Yang Deng et al.

With the aid of large language models, current conversational recommender system (CRS) has gaining strong abilities to persuade users to accept recommended items. While these CRSs are highly persuasive, they can mislead users by incorporating incredible information in their explanations, ultimately damaging the long-term trust between users and the CRS. To address this, we propose a simple yet effective method, called PC-CRS, to enhance the credibility of CRS's explanations during persuasion. It guides the explanation generation through our proposed credibility-aware persuasive strategies and then gradually refines explanations via post-hoc self-reflection. Experimental results demonstrate the efficacy of PC-CRS in promoting persuasive and credible explanations. Further analysis reveals the reason behind current methods producing incredible explanations and the potential of credible explanations to improve recommendation accuracy.

Hongru Liang, Jia Liu, Weihong Du et al.

The machine reading comprehension (MRC) of user manuals has huge potential in customer service. However, current methods have trouble answering complex questions. Therefore, we introduce the Knowing-how & Knowing-that task that requires the model to answer factoid-style, procedure-style, and inconsistent questions about user manuals. We resolve this task by jointly representing the steps and facts in a graph TARA, which supports a unified inference of various questions. Towards a systematical benchmarking study, we design a heuristic method to automatically parse user manuals into TARAs and build an annotated dataset to test the model's ability in answering real-world questions. Empirical results demonstrate that representing user manuals as TARAs is a desired solution for the MRC of user manuals. An in-depth investigation of TARA further sheds light on the issues and broader impacts of future representations of user manuals. We hope our work can move the MRC of user manuals to a more complex and realistic stage.

Mingjie Xu, Jinpeng Chen, Yuzhi Zhao et al.

Multimodal large language models (MLLMs) have enabled a wide range of advanced vision-language applications, including fine-grained object recognition and contextual understanding. When querying specific regions or objects in an image, human users naturally use "visual prompts" (VPs), such as bounding boxes, to provide reference. However, no existing benchmark systematically evaluates the ability of MLLMs to interpret such VPs. This gap leaves it unclear whether current MLLMs can effectively recognize VPs, an intuitive prompting method for humans, and use them to solve problems. To address this limitation, we introduce VP-Bench, a benchmark for assessing MLLMs' capability in VP perception and utilization. VP-Bench employs a two-stage evaluation framework: Stage 1 examines models' ability to perceive VPs in natural scenes, using 30k visualized prompts spanning eight shapes and 355 attribute combinations. Stage 2 investigates the impact of VPs on downstream tasks, measuring their effectiveness in real-world problem-solving scenarios. Using VP-Bench, we evaluate 28 MLLMs, including proprietary systems (e.g., GPT-4o) and open-source models (e.g., InternVL3 and Qwen2.5-VL), and provide a comprehensive analysis of factors that affect VP understanding, such as variations in VP attributes, question arrangement, and model scale. VP-Bench establishes a new reference framework for studying how MLLMs comprehend and resolve grounded referring questions.

Kexin Yang, Dayiheng Liu, Wenqiang Lei et al.

Product description generation is a challenging and under-explored task. Most such work takes a set of product attributes as inputs then generates a description from scratch in a single pass. However, this widespread paradigm might be limited when facing the dynamic wishes of users on constraining the description, such as deleting or adding the content of a user-specified attribute based on the previous version. To address this challenge, we explore a new draft-command-edit manner in description generation, leading to the proposed new task-controllable text editing in E-commerce. More specifically, we allow systems to receive a command (deleting or adding) from the user and then generate a description by flexibly modifying the content based on the previous version. It is easier and more practical to meet the new needs by modifying previous versions than generating from scratch. Furthermore, we design a data augmentation method to remedy the low resource challenge in this task, which contains a model-based and a rule-based strategy to imitate the edit by humans. To accompany this new task, we present a human-written draft-command-edit dataset called E-cEdits and a new metric "Attribute Edit". Our experimental results show that using the new data augmentation method outperforms baselines to a greater extent in both automatic and human evaluations.

Tong Zhang, Peixin Qin, Yang Deng et al.

Large language models (LLMs) are increasingly used to meet user information needs, but their effectiveness in dealing with user queries that contain various types of ambiguity remains unknown, ultimately risking user trust and satisfaction. To this end, we introduce CLAMBER, a benchmark for evaluating LLMs using a well-organized taxonomy. Building upon the taxonomy, we construct ~12K high-quality data to assess the strengths, weaknesses, and potential risks of various off-the-shelf LLMs. Our findings indicate the limited practical utility of current LLMs in identifying and clarifying ambiguous user queries, even enhanced by chain-of-thought (CoT) and few-shot prompting. These techniques may result in overconfidence in LLMs and yield only marginal enhancements in identifying ambiguity. Furthermore, current LLMs fall short in generating high-quality clarifying questions due to a lack of conflict resolution and inaccurate utilization of inherent knowledge. In this paper, CLAMBER presents a guidance and promotes further research on proactive and trustworthy LLMs. Our dataset is available at https://github.com/zt991211/CLAMBER

Weikang Zhang, Zimo Zhu, Zhichuan Yang et al.

Simulating Standardized Patients with cognitive impairment offers a scalable and ethical solution for clinical training. However, existing methods rely on discrete prompt engineering and fail to capture the heterogeneity of deficits across varying domains and severity levels. To address this limitation, we propose StsPatient for the fine-grained simulation of cognitively impaired patients. We innovatively capture domain-specific features by extracting steering vectors from contrastive pairs of instructions and responses. Furthermore, we introduce a Stochastic Token Modulation (STM) mechanism to regulate the intervention probability. STM enables precise control over impairment severity while mitigating the instability of conventional vector methods. Comprehensive experiments demonstrate that StsPatient significantly outperforms baselines in both clinical authenticity and severity controllability.

Qianlong Li, Chen Huang, Shuai Li et al.

Complex Table Question Answering involves providing accurate answers to specific questions based on intricate tables that exhibit complex layouts and flexible header locations. Despite considerable progress having been made in the LLM era, the reasoning processes of existing methods are often implicit, feeding the entire table into prompts, making it difficult to effectively filter out irrelevant information in the table. To this end, we propose GraphOTTER that explicitly establishes the reasoning process to pinpoint the correct answers. In particular, GraphOTTER leverages a graph-based representation, transforming the complex table into an undirected graph. It then conducts step-by-step reasoning on the graph, with each step guided by a set of pre-defined intermediate reasoning actions. As such, it constructs a clear reasoning path and effectively identifies the answer to a given question. Comprehensive experiments on two benchmark datasets and two LLM backbones demonstrate the effectiveness of GraphOTTER. Further analysis indicates that its success may be attributed to the ability to efficiently filter out irrelevant information, thereby focusing the reasoning process on the most pertinent data. Our code and experimental datasets are available at \url{https://github.com/JDing0521/GraphOTTER}.

Duanyu Feng, Li Ding, Hongru Liang et al.

Edge classification, a crucial task for graph applications, remains relatively under-explored compared to link prediction. Current methods often overlook the potential causal influences of node features on edge features, leading to a loss of relevant prior information. In this work, we present an empirical exploration using the Causal Edge Classification Framework (CECF). Unlike conventional causal inference methods, CECF is the first framework to apply causal inference principles to the edge classification task and to explore modeling edge features as a high-dimensional treatment within a causal framework. Based on the node embedding of Graph Neural Network (GNN), CECF seeks to learn a balanced representation of high-dimensional edge features by mitigating the potential influence of node features. Then, a cross-attention network captures the complex dependencies between node and edge features for final edge classification.Extensive experiments demonstrate that CECF not only achieves superior performance but also serves as a flexible, plug-and-play enhancement for existing methods.We also provide empirical analyses, offering insights into when and how this high-dimensional causal modeling framework works for the edge classification.

Youcheng Huang, Fengbin Zhu, Jingkun Tang et al.

Visual Language Models (VLMs) are vulnerable to adversarial attacks, especially those from adversarial images, which is however under-explored in literature. To facilitate research on this critical safety problem, we first construct a new laRge-scale Adervsarial images dataset with Diverse hArmful Responses (RADAR), given that existing datasets are either small-scale or only contain limited types of harmful responses. With the new RADAR dataset, we further develop a novel and effective iN-time Embedding-based AdveRSarial Image DEtection (NEARSIDE) method, which exploits a single vector that distilled from the hidden states of VLMs, which we call the attacking direction, to achieve the detection of adversarial images against benign ones in the input. Extensive experiments with two victim VLMs, LLaVA and MiniGPT-4, well demonstrate the effectiveness, efficiency, and cross-model transferrability of our proposed method. Our code is available at https://github.com/mob-scu/RADAR-NEARSIDE

Li Ding, Duanyu Feng, Chen Huang et al.

Protein-Text Question Answering (QA) is crucial for interpreting biological sequences through natural language. The integration of Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) that efficiently leverages biological databases and facilitates reasoning offers a potent approach for it. However, constrained by the standard RAG pipeline, these models often rely on curated, static datasets instead of expert-proven biological workflows, lacking the fine-grained information processing and struggling to generalize to novel (OOD) proteins. To bridge this gap, we propose 2D-ProteinRAG, a novel framework that empowers LLMs to operate within the gold-standard biological research workflow (BLAST). To further extract high-quality information from noisy retrieval contexts, we introduce a dual-dimensional (2D) filtering strategy following the expert analytical paradigms. Horizontal Fine-grained Attribute Alignment utilizes a lightweight, intent-aware discriminative filter to prune irrelevant metadata and align database entries with specific user queries. Vertical Homology-based Semantic Denoising resolves functional contradictions and redundancy across multiple homologs via hierarchical clustering. Extensive evaluations on both In-Distribution and diverse biological OOD benchmarks demonstrate that 2D-ProteinRAG consistently achieves state-of-the-art performance, outperforming fine-tuned baselines and other RAG methods. Our results validate the framework's robustness and scalability, providing a practical solution for interpreting protein functions in real-world scientific scenarios.

Weihong Du, Wenrui Liao, Hongru Liang et al.

Automatic extraction of procedural graphs from documents creates a low-cost way for users to easily understand a complex procedure by skimming visual graphs. Despite the progress in recent studies, it remains unanswered: whether the existing studies have well solved this task (Q1) and whether the emerging large language models (LLMs) can bring new opportunities to this task (Q2). To this end, we propose a new benchmark PAGED, equipped with a large high-quality dataset and standard evaluations. It investigates five state-of-the-art baselines, revealing that they fail to extract optimal procedural graphs well because of their heavy reliance on hand-written rules and limited available data. We further involve three advanced LLMs in PAGED and enhance them with a novel self-refine strategy. The results point out the advantages of LLMs in identifying textual elements and their gaps in building logical structures. We hope PAGED can serve as a major landmark for automatic procedural graph extraction and the investigations in PAGED can offer insights into the research on logic reasoning among non-sequential elements.

Xinwei Yang, Zhaofeng Liu, Chen Huang et al.

While recent research increasingly emphasizes the value of human-LLM collaboration in competitive programming and proposes numerous empirical methods, a comprehensive understanding remains elusive due to the fragmented nature of existing studies and their use of diverse, application-specific human feedback. Thus, our work serves a three-fold purpose: First, we present the first taxonomy of human feedback consolidating the entire programming process, which promotes fine-grained evaluation. Second, we introduce ELABORATIONSET, a novel programming dataset specifically designed for human-LLM collaboration, meticulously annotated to enable large-scale simulated human feedback and facilitate costeffective real human interaction studies. Third, we introduce ELABORATION, a novel benchmark to facilitate a thorough assessment of human-LLM competitive programming. With ELABORATION, we pinpoint strengthes and weaknesses of existing methods, thereby setting the foundation for future improvement. Our code and dataset are available at https://github.com/SCUNLP/ELABORATION

Shuai Li, Chen Huang, Duanyu Feng et al.

To augment Large Language Models (LLMs) for multi-hop question answering, a mainstream solution within Graph Retrieval Augmented Generation (GraphRAG) leverages lightweight retrievers to efficiently extract information from a given Knowledge Graph (KG). However, existing methods often overlook the inherent challenge of sparse semantic information in graphs. Specifically, our experiments reveal that these methods produce biased retrieval Semantic Shortcut Bias and Long-Tail Path Bias, leading to inadequate semantic modeling and limited GraphRAG effectiveness. To address these issues, we propose STAR, a semantic-tuned and tail-adaptive retriever for GraphRAG. STAR integrates two key learning paradigms: token-level interaction learning and path-weighted contrastive learning. The former employs a cross-attention architecture and a hard path mining mechanism to jointly model the query and path, thereby mitigating the Semantic Shortcut Bias. The latter introduces a tailored contrastive learning objective that utilizes tail-adaptive path weighting, designed to optimize the training process and ease the Long-Tail Path Bias. Extensive experiments demonstrate that STAR consistently outperforms baselines, achieving average retrieval performance gains of 1.8\% and LLM QA performance improvements of 2.2\% across all benchmark datasets. Our code is available at https://anonymous.4open.science/r/STAR-C583.

Weihong Du, Jia Liu, Zujie Wen et al.

It is time-saving to build a reading assistant for customer service representations (CSRs) when reading user manuals, especially information-rich ones. Current solutions don't fit the online custom service scenarios well due to the lack of attention to user questions and possible responses. Hence, we propose to develop a time-saving and careful reading assistant for CSRs, named CARE. It can help the CSRs quickly find proper responses from the user manuals via explicit clue chains. Specifically, each of the clue chains is formed by inferring over the user manuals, starting from the question clue aligned with the user question and ending at a possible response. To overcome the shortage of supervised data, we adopt the self-supervised strategy for model learning. The offline experiment shows that CARE is efficient in automatically inferring accurate responses from the user manual. The online experiment further demonstrates the superiority of CARE to reduce CSRs' reading burden and keep high service quality, in particular with >35% decrease in time spent and keeping a >0.75 ICC score.

Chang Wu, Zhiyuan Liu, Wen Shu et al.

Masked graph modeling (MGM) is a promising approach for molecular representation learning (MRL).However, extending the success of re-mask decoding from 2D to 3D MGM is non-trivial, primarily due to two conflicting challenges: avoiding 2D structure leakage to the decoder, while still providing sufficient 2D context for reconstructing re-masked atoms. To address these challenges, we propose 3D-GSRD: a 3D Molecular Graph Auto-Encoder with Selective Re-mask Decoding. The core innovation of 3D-GSRD lies in its Selective Re-mask Decoding(SRD), which re-masks only 3D-relevant information from encoder representations while preserving the 2D graph structures. This SRD is synergistically integrated with a 3D Relational-Transformer(3D-ReTrans) encoder alongside a structure-independent decoder. We analyze that SRD, combined with the structure-independent decoder, enhances the encoder's role in MRL. Extensive experiments show that 3D-GSRD achieves strong downstream performance, setting a new state-of-the-art on 7 out of 8 targets in the widely used MD17 molecular property prediction benchmark. The code is released at https://github.com/WuChang0124/3D-GSRD.

Jingyao Liu, Chen Huang, Zhizhao Guan et al.

The rapid advancement in large language models (LLMs) has demonstrated significant potential in End-to-End Software Development (E2ESD). However, existing E2ESD benchmarks are limited by coarse-grained requirement specifications and unreliable evaluation protocols, hindering a true understanding of current framework capabilities. To address these limitations, we present E2EDev, a novel benchmark grounded in the principles of Behavior-Driven Development (BDD), which evaluates the capabilities of E2ESD frameworks by assessing whether the generated software meets user needs through mimicking real user interactions (Figure 1). E2EDev comprises (i) a fine-grained set of user requirements, (ii) multiple BDD test scenarios with corresponding Python step implementations for each requirement, and (iii) a fully automated testing pipeline built on the Behave framework. To ensure its quality while reducing the annotation effort, E2EDev leverages our proposed Human-in-the-Loop Multi-Agent Annotation Framework (HITL-MAA). By evaluating various E2ESD frameworks and LLM backbones with E2EDev, our analysis reveals a persistent struggle to effectively solve these tasks, underscoring the critical need for more effective and cost-efficient E2ESD solutions. Our codebase and benchmark are publicly available at https://github.com/SCUNLP/E2EDev.

Yongjie Xiao, Hongru Liang, Peixin Qin et al.

Despite the great potential of large language models(LLMs) in machine comprehension, it is still disturbing to fully count on them in real-world scenarios. This is probably because there is no rational explanation for whether the comprehension process of LLMs is aligned with that of experts. In this paper, we propose SCOP to carefully examine how LLMs perform during the comprehension process from a cognitive view. Specifically, it is equipped with a systematical definition of five requisite skills during the comprehension process, a strict framework to construct testing data for these skills, and a detailed analysis of advanced open-sourced and closed-sourced LLMs using the testing data. With SCOP, we find that it is still challenging for LLMs to perform an expert-level comprehension process. Even so, we notice that LLMs share some similarities with experts, e.g., performing better at comprehending local information than global information. Further analysis reveals that LLMs can be somewhat unreliable -- they might reach correct answers through flawed comprehension processes. Based on SCOP, we suggest that one direction for improving LLMs is to focus more on the comprehension process, ensuring all comprehension skills are thoroughly developed during training.

Chen Huang, Junkai Luo, Xinzuo Wang et al.

The massive user-generated content (UGC) available in Chinese social media is giving rise to the possibility of studying internet buzzwords. In this paper, we study if large language models (LLMs) can generate accurate definitions for these buzzwords based on UGC as examples. Our work serves a threefold contribution. First, we introduce CHEER, the first dataset of Chinese internet buzzwords, each annotated with a definition and relevant UGC. Second, we propose a novel method, called RESS, to effectively steer the comprehending process of LLMs to produce more accurate buzzword definitions, mirroring the skills of human language learning. Third, with CHEER, we benchmark the strengths and weaknesses of various off-the-shelf definition generation methods and our RESS. Our benchmark demonstrates the effectiveness of RESS while revealing crucial shared challenges: over-reliance on prior exposure, underdeveloped inferential abilities, and difficulty identifying high-quality UGC to facilitate comprehension. We believe our work lays the groundwork for future advancements in LLM-based definition generation. Our dataset and code are available at https://github.com/SCUNLP/Buzzword.

Youcheng Huang, Jingkun Tang, Duanyu Feng et al.

People tell lies when seeking rewards. Large language models (LLMs) are aligned to human values with reinforcement learning where they get rewards if they satisfy human preference. We find that this also induces dishonesty in helpful and harmless alignment where LLMs tell lies in generating harmless responses. Using the latest interpreting tools, we detect dishonesty, show how LLMs can be harmful if their honesty is increased, and analyze such conflicts at the parameter-level. Given these preliminaries and the hypothesis that reward-seeking stimulates dishonesty, we theoretically show that the dishonesty can in-turn decrease the alignment performances and augment reward-seeking alignment with representation regularization. Extensive results, including GPT-4 annotated win-rates, perplexities, and cases studies demonstrate that we can train more honest, helpful, and harmless LLMs. We will make all our codes and results be open-sourced upon this paper's acceptance.

Ruibo Deng, Duanyu Feng, Wenqiang Lei

Offline preference optimization offers a simpler and more stable alternative to RLHF for aligning language models. However, their effectiveness is critically dependent on ranking accuracy, a metric where further gains are highly impactful. This limitation arises from a fundamental problem that we identify and formalize as the Overfitting-Underfitting Dilemma: current margin designs cause models to apply excessive, wasteful gradients to correctly ranked samples (overfitting) while providing insufficient corrective signals for misranked ones (underfitting). To resolve this dilemma, we propose Adaptive Margin-attached Preference Optimization (AMaPO), a simple yet principled algorithm. AMaPO employs an instance-wise adaptive margin, refined by Z-normalization and exponential scaling, which dynamically reallocates learning effort by amplifying gradients for misranked samples and suppressing them for correct ones. Extensive experiments on widely used benchmarks demonstrate that AMaPO not only achieves better ranking accuracy and superior downstream alignment performance, but targeted analysis also confirms that it successfully mitigates the core overfitting and underfitting issues.

Duanyu Feng, Bowen Qin, Chen Huang et al.

Direct Preference Optimization (DPO), which derives reward signals directly from pairwise preference data, has shown its effectiveness on aligning Large Language Models (LLMs) with human preferences. Despite its widespread use across various tasks, DPO has been criticized for its sensitivity to the SFT's effectiveness and its hindrance to the learning capacity towards human-preferred responses, leading to less satisfactory performance. To overcome those limitations, the theoretical understanding of DPO are indispensable but still lacking. To this end, we take a step towards theoretically analyzing and understanding the limitations of DPO. Specifically, we provide an analytical framework using the field theory to analyze the optimization process of DPO. By analyzing the gradient vector field of the DPO loss function, we find that the DPO loss function decreases the probability of producing human dispreferred data at a faster rate than it increases the probability of producing preferred data. This provides theoretical insights for understanding the limitations of DPO discovered in the related research experiments, thereby setting the foundation for its improvement.

Tong Zhang, Chen Huang, Yang Deng et al.

We investigate non-collaborative dialogue agents, which are expected to engage in strategic conversations with diverse users, for securing a mutual agreement that leans favorably towards the system's objectives. This poses two main challenges for existing dialogue agents: 1) The inability to integrate user-specific characteristics into the strategic planning, and 2) The difficulty of training strategic planners that can be generalized to diverse users. To address these challenges, we propose Trip to enhance the capability in tailored strategic planning, incorporating a user-aware strategic planning module and a population-based training paradigm. Through experiments on benchmark non-collaborative dialogue tasks, we demonstrate the effectiveness of Trip in catering to diverse users.

Yue Chen, Chen Huang, Yang Deng et al.

Equipping a conversational search engine with strategies regarding when to ask clarification questions is becoming increasingly important across various domains. Attributing to the context understanding capability of LLMs and their access to domain-specific sources of knowledge, LLM-based clarification strategies feature rapid transfer to various domains in a post-hoc manner. However, they still struggle to deliver promising performance on unseen domains, struggling to achieve effective domain transferability. We take the first step to investigate this issue and existing methods tend to produce one-size-fits-all strategies across diverse domains, limiting their search effectiveness. In response, we introduce a novel method, called Style, to achieve effective domain transferability. Our experimental results indicate that Style bears strong domain transferability, resulting in an average search performance improvement of ~10% on four unseen domains.

Chen Huang, Peixin Qin, Yang Deng et al.

The conversational recommendation system (CRS) has been criticized regarding its user experience in real-world scenarios, despite recent significant progress achieved in academia. Existing evaluation protocols for CRS may prioritize system-centric factors such as effectiveness and fluency in conversation while neglecting user-centric aspects. Thus, we propose a new and inclusive evaluation protocol, Concept, which integrates both system- and user-centric factors. We conceptualise three key characteristics in representing such factors and further divide them into six primary abilities. To implement Concept, we adopt a LLM-based user simulator and evaluator with scoring rubrics that are tailored for each primary ability. Our protocol, Concept, serves a dual purpose. First, it provides an overview of the pros and cons in current CRS models. Second, it pinpoints the problem of low usability in the "omnipotent" ChatGPT and offers a comprehensive reference guide for evaluating CRS, thereby setting the foundation for CRS improvement.

Youcheng Huang, Wenqiang Lei, Zheng Zhang et al.

Knowledge-editing updates knowledge of large language models (LLMs) and contributes to the interpretability and application of LLMs. However, knowledge applying is context-consistent: LLMs can recall the same knowledge in different contexts. Existing works ignore this property and the editing lacks generalization. In this paper, we empirically find that the effects of different contexts upon LLMs in recalling the same knowledge follow a Gaussian-like distribution. We then sample Gaussian noises to simulate the effects of different contexts when updating LLMs. By such, we can make LLMs see the unseen contexts where the edited knowledge will be applied, therefore improving the editing generalization. Experimental results on three LLMs demonstrate the effectiveness of our methods and also distinguish our methods from the others of fine-tuning LLMs by noises.

Wei Ji, Li Li, Zheqi Lv et al.

In our increasingly interconnected world, where intelligent devices continually amass copious personalized multi-modal data, a pressing need arises to deliver high-quality, personalized device-aware services. However, this endeavor presents a multifaceted challenge to prevailing artificial intelligence (AI) systems primarily rooted in the cloud. As these systems grapple with shifting data distributions between the cloud and devices, the traditional approach of fine-tuning-based adaptation (FTA) exists the following issues: the costly and time-consuming data annotation required by FTA and the looming risk of model overfitting. To surmount these challenges, we introduce a Universal On-Device Multi-modal Model Adaptation Framework, revolutionizing on-device model adaptation by striking a balance between efficiency and effectiveness. The framework features the Fast Domain Adaptor (FDA) hosted in the cloud, providing tailored parameters for the Lightweight Multi-modal Model on devices. To enhance adaptability across multi-modal tasks, the AnchorFrame Distribution Reasoner (ADR) minimizes communication costs. Our contributions, encapsulated in the Cloud-Device Collaboration Multi-modal Parameter Generation (CDC-MMPG) framework, represent a pioneering solution for on-Device Multi-modal Model Adaptation (DMMA). Extensive experiments validate the efficiency and effectiveness of our method, particularly in video question answering and retrieval tasks, driving forward the integration of intelligent devices into our daily lives.

Chen Huang, Yang Deng, Wenqiang Lei et al.

To obtain high-quality annotations under limited budget, semi-automatic annotation methods are commonly used, where a portion of the data is annotated by experts and a model is then trained to complete the annotations for the remaining data. However, these methods mainly focus on selecting informative data for expert annotations to improve the model predictive ability (i.e., triage-to-human data), while the rest of the data is indiscriminately assigned to model annotation (i.e., triage-to-model data). This may lead to inefficiencies in budget allocation for annotations, as easy data that the model could accurately annotate may be unnecessarily assigned to the expert, and hard data may be misclassified by the model. As a result, the overall annotation quality may be compromised. To address this issue, we propose a selective annotation framework called SANT. It effectively takes advantage of both the triage-to-human and triage-to-model data through the proposed error-aware triage and bi-weighting mechanisms. As such, informative or hard data is assigned to the expert for annotation, while easy data is handled by the model. Experimental results show that SANT consistently outperforms other baselines, leading to higher-quality annotation through its proper allocation of data to both expert and model workers. We provide pioneering work on data annotation within budget constraints, establishing a landmark for future triage-based annotation studies.

Youcheng Huang, Chen Huang, Duanyu Feng et al.

Understanding the inner workings of Large Language Models (LLMs) is a critical research frontier. Prior research has shown that a single LLM's concept representations can be captured as steering vectors (SVs), enabling the control of LLM behavior (e.g., towards generating harmful content). Our work takes a novel approach by exploring the intricate relationships between concept representations across different LLMs, drawing an intriguing parallel to Plato's Allegory of the Cave. In particular, we introduce a linear transformation method to bridge these representations and present three key findings: 1) Concept representations across different LLMs can be effectively aligned using simple linear transformations, enabling efficient cross-model transfer and behavioral control via SVs. 2) This linear transformation generalizes across concepts, facilitating alignment and control of SVs representing different concepts across LLMs. 3) A weak-to-strong transferability exists between LLM concept representations, whereby SVs extracted from smaller LLMs can effectively control the behavior of larger LLMs.

Chen Huang, Yiping Jin, Ilija Ilievski et al.

Human annotation is a time-consuming task that requires a significant amount of effort. To address this issue, interactive data annotation utilizes an annotation model to provide suggestions for humans to approve or correct. However, annotation models trained with limited labeled data are prone to generating incorrect suggestions, leading to extra human correction effort. To tackle this challenge, we propose Araida, an analogical reasoning-based approach that enhances automatic annotation accuracy in the interactive data annotation setting and reduces the need for human corrections. Araida involves an error-aware integration strategy that dynamically coordinates an annotation model and a k-nearest neighbors (KNN) model, giving more importance to KNN's predictions when predictions from the annotation model are deemed inaccurate. Empirical studies demonstrate that Araida is adaptable to different annotation tasks and models. On average, it reduces human correction labor by 11.02% compared to vanilla interactive data annotation methods.

Yuanheng Fang, Guoqing Chao, Wenqiang Lei et al.

Large Language Models (LLMs) have recently achieved impressive results in complex reasoning tasks through Chain of Thought (CoT) prompting. However, most existing CoT methods rely on using the same prompts, whether manually designed or automatically generated, to handle the entire dataset. This one-size-fits-all approach may fail to meet the specific needs arising from the diversities within a single dataset. To solve this problem, we propose the Clustered Distance-Weighted Chain of Thought (CDW-CoT) method, which dynamically constructs prompts tailored to the characteristics of each data instance by integrating clustering and prompt optimization techniques. Our method employs clustering algorithms to categorize the dataset into distinct groups, from which a candidate pool of prompts is selected to reflect the inherent diversity within the dataset. For each cluster, CDW-CoT trains the optimal prompt probability distribution tailored to their specific characteristics. Finally, it dynamically constructs a unique prompt probability distribution for each test instance, based on its proximity to cluster centers, from which prompts are selected for reasoning. CDW-CoT consistently outperforms traditional CoT methods across six datasets, including commonsense, symbolic, and mathematical reasoning tasks. Specifically, when compared to manual CoT, CDW-CoT achieves an average accuracy improvement of 25.34% on LLaMA2 (13B) and 15.72% on LLaMA3 (8B).

Chen Huang, Yang Deng, Wenqiang Lei et al.

With the advancement of large language models (LLMs), intelligent models have evolved from mere tools to autonomous agents with their own goals and strategies for cooperating with humans. This evolution has birthed a novel paradigm in NLP, i.e., human-model cooperation, that has yielded remarkable progress in numerous NLP tasks in recent years. In this paper, we take the first step to present a thorough review of human-model cooperation, exploring its principles, formalizations, and open challenges. In particular, we introduce a new taxonomy that provides a unified perspective to summarize existing approaches. Also, we discuss potential frontier areas and their corresponding challenges. We regard our work as an entry point, paving the way for more breakthrough research in this regard.

Chen Huang, Peixin Qin, Wenqiang Lei et al.

One of the key factors in language productivity and human cognition is the ability of systematic compositionality, which refers to understanding composed unseen examples of seen primitives. However, recent evidence reveals that the Transformers have difficulty generalizing the composed context based on the seen primitives. To this end, we take the first step to propose a compositionality-aware Transformer called CAT and two novel pre-training tasks to facilitate systematic compositionality. We tentatively provide a successful implementation of a multi-layer CAT on the basis of the especially popular BERT. The experimental results demonstrate that CAT outperforms baselines on compositionality-aware tasks with minimal impact on the effectiveness on standardized language understanding tasks.

Chen Huang, Duanyu Feng, Wenqiang Lei et al.

Deploying dense retrieval models efficiently is becoming increasingly important across various industries. This is especially true for enterprise search services, where customizing search engines to meet the time demands of different enterprises in different domains is crucial. Motivated by this, we develop a time-efficient approach called DREditor to edit the matching rule of an off-the-shelf dense retrieval model to suit a specific domain. This is achieved by directly calibrating the output embeddings of the model using an efficient and effective linear mapping. This mapping is powered by an edit operator that is obtained by solving a specially constructed least squares problem. Compared to implicit rule modification via long-time finetuning, our experimental results show that DREditor provides significant advantages on different domain-specific datasets, dataset sources, retrieval models, and computing devices. It consistently enhances time efficiency by 100-300 times while maintaining comparable or even superior retrieval performance. In a broader context, we take the first step to introduce a novel embedding calibration approach for the retrieval task, filling the technical blank in the current field of embedding calibration. This approach also paves the way for building domain-specific dense retrieval models efficiently and inexpensively.