论文检索

Xuan Li, Zhe Yin, Xiaodong Gu, Beijun Shen

With the widespread use of LLMs, preserving privacy in user prompts has become crucial, as prompts risk exposing private and sensitive data to cloud LLMs. Conventional techniques like homomorphic encryption (HE), secure multi-party computation, and federated learning (FL) are not well-suited to this scenario due to the lack of control over user participation in remote model interactions. In this paper, we propose PromptObfus, a novel method for desensitizing LLM prompts. The core idea of PromptObfus is "anti-adversarial" learning, which perturbs sensitive words in the prompt to obscure private information while retaining the stability of model predictions. Specifically, PromptObfus frames prompt desensitization as a masked language modeling task, replacing privacy-sensitive terms with a [MASK] token. A desensitization model is utilized to generate candidate replacements for each masked position. These candidates are subsequently selected based on gradient feedback from a surrogate model, ensuring minimal disruption to the task output. We demonstrate the effectiveness of our approach on three NLP tasks. Results show that PromptObfus effectively prevents privacy inference from remote LLMs while preserving task performance.

Xinran Li, Yu Liu, Jiaqi Qiao, Xiujuan Xu

Emotion Recognition in Conversation (ERC) is a crucial task for understanding human emotions and enabling natural human-computer interaction. Although Large Language Models (LLMs) have recently shown great potential in this field, their ability to capture the intrinsic connections between explicit and implicit emotions remains limited. We propose a novel ERC training framework, PRC-Emo, which integrates Prompt engineering, demonstration Retrieval, and Curriculum learning, with the goal of exploring whether LLMs can effectively perceive emotions in conversational contexts. Specifically, we design emotion-sensitive prompt templates based on both explicit and implicit emotional cues to better guide the model in understanding the speaker’s psychological states. We construct the first dedicated demonstration retrieval repository for ERC, which includes training samples from widely used datasets, as well as high-quality dialogue examples generated by LLMs and manually verified. Moreover, we introduce a curriculum learning strategy into the LoRA fine-tuning process, incorporating weighted emotional shifts between same-speaker and different-speaker utterances to assign difficulty levels to dialogue samples, which are then organized in an easy-to-hard training sequence. Experimental results on two benchmark datasets—IEMOCAP and MELD—show that our method achieves new state-of-the-art (SOTA) performance, demonstrating the effectiveness and generalizability of our approach in improving LLM-based emotional understanding.

Xingyu Li, Xiaolei Liu, Cheng Liu, Yixiao Xu, Kangyi Ding, Bangzhou Xin, Jia-Li Yin

As large language models (LLMs) scale, their inference incurs substantial computational resources, exposing them to energy-latency attacks, where crafted prompts induce high energy and latency cost. Existing attack methods aim to prolong output by delaying the generation of termination symbols. However, as the output grows longer, controlling the termination symbols through input becomes difficult, making these methods less effective. Therefore, we propose LoopLLM, an energy-latency attack framework based on the observation that repetitive generation can trigger low-entropy decoding loops, reliably compelling LLMs to generate until their output limits. LoopLLM introduces (1) a repetition-inducing prompt optimization that exploits autoregressive vulnerabilities to induce repetitive generation, and (2) a token-aligned ensemble optimization that aggregates gradients to improve cross-model transferability. Extensive experiments on 12 open-source and 2 commercial LLMs show that LoopLLM significantly outperforms existing methods, achieving over 90% of the maximum output length, compared to 20% for baselines, and improving transferability by around 40% to DeepSeek-V3 and Gemini 2.5 Flash.

Xiaomin Li, Mingye Gao, Zhiwei Zhang, Chang Yue, Hong Hu

High-quality training data is critical to the performance of large language models (LLMs). Recent work has explored using LLMs to rate and select data based on a small set of human-designed criteria (rules), but these approaches often rely heavily on heuristics, lack principled metrics for rule evaluation, and generalize poorly to new tasks. We propose a novel rule-based data selection framework that introduces a metric based on the orthogonality of rule score vectors to evaluate and select complementary rules. Our automated pipeline first uses LLMs to generate diverse rules covering multiple aspects of data quality, then rates samples according to these rules and applies the determinantal point process (DPP) to select the most independent rules. These rules are then used to score the full dataset, and high-scoring samples are selected for downstream tasks such as LLM fine-tuning. We evaluate our framework in two experiment setups: (1) alignment with ground-truth ratings and (2) performance of LLMs fine-tuned on the selected data. Experiments across IMDB, Medical, Math, and Code domains demonstrate that our DPP-based rule selection consistently improves both rating accuracy and downstream model performance over strong baselines.

Xiaomin Li, Zhou Yu, Ziji Zhang, Yingying Zhuang, Swair Shah, Narayanan Sadagopan, Anurag Beniwal

Large language models (LLMs) have demonstrated remarkable performance across diverse tasks by encoding vast amounts of factual knowledge. However, they are still prone to hallucinations, generating incorrect or misleading information, often accompanied by high uncertainty. Existing methods for hallucination detection primarily focus on quantifying internal uncertainty, which arises from missing or conflicting knowledge within the model. However, hallucinations can also stem from external uncertainty, where ambiguous user queries lead to multiple possible interpretations. In this work, we introduce **Semantic Volume**, a novel mathematical measure for quantifying both external and internal uncertainty in LLMs. Our approach perturbs queries and responses, embeds them in a semantic space, and computes the Gram matrix determinant of the embedding vectors, capturing their dispersion as a measure of uncertainty. Our framework provides a generalizable and unsupervised uncertainty detection method without requiring internal access to LLMs. We conduct extensive experiments on both external and internal uncertainty detections, demonstrating that our Semantic Volume method consistently outperforms existing baselines in both tasks. Additionally, we provide theoretical insights linking our measure to differential entropy, unifying and extending previous sampling-based uncertainty measures such as the semantic entropy. Semantic Volume is shown to be a robust and interpretable approach to improving the reliability of LLMs by systematically detecting uncertainty in both user queries and model responses.

Xiaomin Li, Xupeng Chen, Jingxuan Fan, Eric Hanchen Jiang, Mingye Gao

The safety alignment of large language models (LLMs) often relies on reinforcement learning from human feedback (RLHF), which requires human annotations to construct preference datasets. Given the challenge of assigning overall quality scores to data, recent works increasingly adopt fine-grained ratings based on multiple safety rules. In this paper, we discover a robust phenomenon: Rules with higher rating entropy tend to have lower accuracy in distinguishing human-preferred responses. Exploiting this insight, we propose ENCORE, a simple entropy-guided method to compose multi-head rewards by penalizing rules with high rating entropy. Theoretically, we show that such rules yield negligible weights under the Bradley–Terry loss during weight optimization, naturally justifying their penalization. Empirically, ENCORE consistently outperforms strong baselines, including random and uniform weighting, single-head Bradley–Terry, and LLM-as-a-judge, etc. on RewardBench safety tasks. Our method is completely training-free, generally applicable across datasets, and retains interpretability, making it a practical and effective approach for multi-attribute reward modeling.

Weijie Li, Jin Wang, Liang-Chih Yu, Xuejie Zhang

Large reasoning models (LRMs) improve performance at test time by thinking longer, but this often leads to overthinking and high computational cost. To address this, recent reinforcement learning (RL) methods adopt outcome-level rewards, such as rule- or prompt-based signals, that favor shorter correct reasoning paths but often overlook reasoning quality. While such rewards neglect intermediate reasoning, dense supervision from process reward models (PRMs) has proven more effective in promoting coherent and high-quality reasoning. However, static PRM supervision introduces two challenges: reward hacking, since fixed rewards poorly capture global reasoning objectives, and the high training cost of obtaining dense reward labels at scale. To overcome these issues, we propose Step Group Relative Policy Optimization (Step-GRPO), a GRPO-based method that integrates step-level PRM signals into sparse trajectory-level feedback, avoiding costly step-level supervision while improving reasoning quality beyond accuracy. In addition, Step-GRPO employs a step-attention mechanism that captures inter-step dependencies and emphasizes critical reasoning steps, effectively mitigating reward hacking. We apply Step-GRPO to train large language models and observe consistent gains in reasoning quality, accuracy, and shorter reasoning traces across multiple math benchmarks, outperforming reinforcement learning baselines at substantially lower cost. Notably, the proposed model achieves 36.7 percent accuracy on AIME 2024 with 11,000 training samples and a training cost of 38 US dollars, surpassing baselines that require over 1,000 US dollars and more than 40,000 samples, demonstrating strong cost-effectiveness and scalability.

Shuocheng Li, Yihao Liu, Silin Du, Wenxuan Zeng, Zhe Xu, Mengyu Zhou, Yeye He, Haoyu Dong, Shi Han, Dongmei Zhang

Large language models (LLMs) have shown great promise in automating data science workflows. However, existing models still struggle with multi-step reasoning and tool use, limiting their effectiveness on complex data analysis tasks. To address this limitation, we propose a scalable pipeline that extracts high-quality, tool-based data analysis tasks and their executable multi-step solutions from real-world Jupyter notebooks and associated data files. Using this pipeline, we introduce NbQA, a large-scale dataset of standardized task–solution pairs that reflect authentic tool-use patterns in practical data science scenarios. To further enhance the multi-step reasoning capabilities, we present Jupiter, a framework that formulates data analysis as a search problem and applies Monte Carlo Tree Search (MCTS) to generate diverse solution trajectories for value model learning. During inference, Jupiter combines the value model and node visit counts to efficiently collect executable multi-step plans with minimal search steps. Experimental results show that Qwen2.5-7B and 14B-Instruct models on NbQA solve 77.82% and 86.38% of tasks on InfiAgent-DABench, respectively—matching or surpassing GPT-4o and advanced agent frameworks. Further evaluations demonstrate improved generalization and stronger tool-use reasoning across diverse multi-step reasoning tasks.

Shuaimin Li, Liyang Fan, Yufang Lin, Zeyang Li, Xian Wei, Shiwen Ni, Hamid Alinejad-Rokny, Min Yang

Existing paper review methods often rely on superficial manuscript features or directly on large language models (LLMs), which are prone to hallucinations, biased scoring, and limited reasoning capabilities. Moreover, these methods often fail to capture the complex argumentative reasoning and negotiation dynamics inherent in reviewer-author interactions. To address these limitations, we propose ReViewGraph (Reviewer-Author Debates Graph Reasoner), a novel framework that performs heterogeneous graph reasoning over LLM-simulated multi-round reviewer-author debates. In our approach, reviewer-author exchanges are simulated through LLM-based multi-agent collaboration. Diverse opinion relations (e.g., acceptance, rejection, clarification, and compromise) are then explicitly extracted and encoded as typed edges within a heterogeneous interaction graph. By applying graph neural networks to reason over these structured debate graphs, ReViewGraph captures fine-grained argumentative dynamics and enables more informed review decisions. Extensive experiments on three datasets demonstrate that ReViewGraph outperforms strong baselines with an average relative improvement of 15.73%, underscoring the value of modeling detailed reviewer–author debate structures.

Shike Li, Xiaokai Wang, Xiaofeng Liu, Xin Tong, Hu Zhang

Large language models (LLMs) frequently generate fluent yet factually inaccurate content, a phenomenon known as hallucination. Recent inference-time approaches aim to improve truthfulness by steering model activations toward semantically meaningful directions. While effective to some extent, these methods typically process activations independently, neglecting the internal coordination structure of multi-head attention (MHA), where attention heads interact to form semantic representations. In this work, we propose CoFact, an adaptive inference-time mechanism that improves factual consistency by dynamically coordinating attention head behaviors. Inspired by cooperative game theory, CoFact conceptualizes attention heads as collaborative agents. It models the semantic utility and redundancy of each head and adaptively modulates their contributions to the final attention output. Notably, rather than directly altering intermediate representations, CoFact performs token-level coordination to encourage diverse and complementary attention patterns across heads. CoFact is plug-and-play compatible with mainstream LLM architectures and requires no additional supervision or model retraining. Experimental results across multiple standard factuality benchmarks demonstrate that CoFact consistently enhances factual accuracy while maintaining generation fluency.

Shangyu Li, Juyong Jiang, Tiancheng Zhao, Jiasi Shen

We introduce OSVBench, a new benchmark for evaluating Large Language Models (LLMs) on the task of generating complete formal specifications for verifying the functional correctness of operating system kernels. This benchmark is built upon a real-world operating system kernel, Hyperkernel, and consists of 245 complex specification generation tasks in total, each of which is a long-context task of about 20k-30k tokens. The benchmark formulates the specification generation task as a program synthesis problem confined to a domain for specifying states and transitions. This formulation is provided to LLMs through a programming model. The LLMs must be able to understand the programming model and verification assumptions before delineating the correct search space for syntax and semantics and generating formal specifications. Guided by the operating system's high-level functional description, the LLMs are asked to generate a specification that fully describes all correct states and transitions for a potentially buggy code implementation of the operating system. Experimental results with 12 state-of-the-art LLMs indicate limited performance of existing LLMs on the specification generation task for operating system verification. Significant disparities in their performance highlight differences in their ability to handle long-context code generation tasks.

Qizhi Li, Xuyang Wang, Yingke Chen, Ming Yan, Dezhong Peng, Xi Peng, Xu Wang

Recent advancements in Pre-trained Language Models (PLMs) have significantly enhanced performance across various Natural Language Processing (NLP) tasks. However, the variability in data distributions across different domains presents challenges in generalizing these models to unseen domains. Domain generalization offers a promising solution, but existing text domain generalization methods typically rely on adversarial training to learn domain-invariant features, which often leads to models with high computational and memory overhead. To address this issue, this paper proposes a novel solution named Generalization via Prompts and Contrastive Learning (GenPromptCL) to enhance the generalization capability in domain generalization. GenPromptCL consists of two key components: Domain-Misleading Prompt Learning (DMPL) and Pseudo Label-based Contrastive Learning (PCL). Specifically, DMPL disrupts domain labels randomly, misleading the model into producing incorrect domain labels. This forces the model to learn domain-invariant features. Meanwhile, PCL generates pseudo labels within a single mini-batch, enabling the model to learn both intra-class and inter-class discriminative representations with low time and space complexity. Extensive experimental results demonstrate that GenPromptCL achieves state-of-the-art performance on three distinct text classification tasks (sentiment analysis, rumor detection, and natural language inference) while significantly improving model operation efficiency.

Qiyang Li, Rui Kong, Yuchen Li, Hengyi Cai, Shuaiqiang Wang, Linghe Kong, Guihai Chen, Dawei Yin

The integration of dynamic, sparse structures like Mixture-of-Experts (MoE) with parameter-efficient adapters (e.g., LoRA) is a powerful technique for enhancing Large Language Models (LLMs). However, this architectural enhancement comes at a steep cost: despite minimal increases in computational load, the inference latency often skyrockets, leading to decoding speeds slowing by over 2.5 times. Through a fine-grained performance analysis, we pinpoint the primary bottleneck not in the computation itself, but in the severe overhead from fragmented, sequential CUDA kernel launches required for conventional dynamic routing. To address this challenge, we introduce AdaFuse, a framework built on a tight co-design between the algorithm and the underlying hardware system to enable efficient dynamic adapter execution. Departing from conventional layer-wise or block-wise routing, AdaFuse employs a token-level pre-gating strategy, which makes a single, global routing decision for all adapter layers before a token is processed. This ``decide-once, apply-everywhere'' approach effectively staticizes the execution path for each token, creating an opportunity for holistic optimization. We capitalize on this by developing a custom CUDA kernel that performs a fused switching operation, merging the parameters of all selected LoRA adapters into the backbone model in a single, efficient pass. Experimental results on popular open-source LLMs show that AdaFuse achieves accuracy on par with state-of-the-art dynamic adapters while drastically cutting decoding latency by a factor of over 2.4x, thereby bridging the gap between model capability and inference efficiency.

Qingchuan Li, Mingyue Cheng, Zirui Liu, Daoyu Wang, Yuting Zeng, Tongxuan Liu

Logical reasoning is a core challenge in natural language understanding and a fundamental capability of artificial intelligence, underpinning scientific discovery, mathematical theorem proving, and complex decision-making. Despite the remarkable progress of large language models (LLMs), most current approaches still rely on forward reasoning paradigms, generating step-by-step rationales from premises to conclusions. However, such methods often suffer from redundant inference paths, hallucinated steps, and semantic drift, resulting in inefficient and unreliable reasoning. In this paper, we propose a novel framework, Hypothesis-driven Backward Logical Reasoning (HBLR). The core idea is to integrate confidence-aware symbolic translation with hypothesis-driven backward reasoning. In the translation phase, only high-confidence spans are converted into logical form, such as first-order logic (FOL), while uncertain content remains in natural language. A translation reflection module further ensures semantic fidelity by evaluating symbolic outputs and reverting lossy ones back to text when necessary. In the reasoning phase, HBLR simulates human deductive thinking by assuming the conclusion is true and recursively verifying its premises. A reasoning reflection module further identifies and corrects flawed inference steps, enhancing logical coherence. Extensive experiments on five reasoning benchmarks demonstrate that HBLR consistently outperforms strong baselines in both accuracy and efficiency.

Qinfeng Li, Miao Pan, Jintao Chen, Fu Teng, Zhiqiang Shen, Ge Su, Hao Peng, Xuhong Zhang

Model merging has emerged as an efficient technique for expanding large language models (LLMs) by integrating specialized expert models. However, it also introduces a new threat: model merging stealing, where free-riders exploit models through unauthorized model merging. Unfortunately, existing defense mechanisms fail to provide effective protection. Specifically, we identify three critical protection properties that existing methods fail to simultaneously satisfy: (1) proactively preventing unauthorized merging; (2) ensuring compatibility with general open-source settings; (3) achieving high security with negligible performance loss. To address the above issues, we propose MergeBarrier, a plug-and-play defense that proactively prevents unauthorized merging. The core design of MergeBarrier is to disrupt the Linear Mode Connectivity (LMC) between the protected model and its homologous counterparts, thereby eliminating the low-loss path required for effective model merging. Extensive experiments show that MergeBarrier effectively prevents model merging stealing with negligible accuracy loss.

Qinfeng Li, Miao Pan, Ke Xiong, Ge Su, Zhiqiang Shen, Yan Liu, Sun Bing, Hao Peng, Xuhong Zhang

Retrieval-Augmented Generation (RAG) systems deployed over proprietary knowledge bases face growing threats from reconstruction attacks that aggregate model responses to replicate knowledge bases. Such attacks exploit both intra-class and inter-class paths—progressively extracting fine-grained knowledge within topics and diffusing it across semantically related ones, thereby enabling comprehensive extraction of the original knowledge base. However, existing defenses target only one path, leaving the other unprotected. We conduct a systematic exploration to assess the impact of protecting each path independently and find that joint protection is essential for effective defense. Based on this, we propose RAGFort, a structure-aware dual-module defense combining contrastive reindexing for inter-class isolation and constrained cascade generation for intra-class protection. Experiments across security, performance, and robustness confirm that RAGFort significantly reduces reconstruction success while preserving answer quality, offering the first comprehensive defense against knowledge base extraction attacks.

Qiming Li, Zekai Ye, Xiaocheng Feng, Weihong Zhong, Weitao Ma, Xiachong Feng

Despite the remarkable advancements of Large Vision-Language Models (LVLMs), the mechanistic interpretability remains underexplored. Existing analyses are insufficiently comprehensive and lack examination covering visual and textual tokens, model components, and the full range of layers. This limitation restricts actionable insights to improve the faithfulness of model output and the development of downstream tasks, such as hallucination mitigation. To address this limitation, we introduce Fine-grained Cross-modal Causal Tracing (FCCT) framework, which systematically quantifies the causal effects on visual object perception. FCCT conducts fine-grained analysis covering the full range of visual and textual tokens, three core model components including multi-head self-attention (MHSA), feed-forward networks (FFNs), and hidden states, across all decoder layers. Our analysis is the first to demonstrate that MHSAs of the last token in middle layers play a critical role in aggregating cross-modal information, while FFNs exhibit a three-stage hierarchical progression for the storage and transfer of visual object representations. Building on these insights, we propose Intermediate Representation Injection (IRI), a training-free inference-time technique that reinforces visual object information flow by precisely intervening on cross-modal representations at specific components and layers, thereby enhancing perception and mitigating hallucination. Consistent improvements across five widely used benchmarks and LVLMs demonstrate IRI achieves state-of-the-art performance, while preserving inference speed and other foundational performance.

Mengfan Li, Xuanhua Shi, Yang Deng

Large Language models (LLMs) are revolutionizing the conversational recommender systems (CRS) through their impressive capabilities in instruction comprehension, reasoning, and human interaction. A core factor underlying effective dialogue is the ability to infer and reason about others' mental states (such as desire, intention, and belief), a cognitive capacity commonly referred to as Theory of Mind (ToM). Despite growing interest in evaluating ToM in LLMs, current benchmarks predominantly rely on synthetic narratives inspired by Sally-Anne test, which emphasize physical perception and fail to capture the complexity of mental state inference in real-world conversational settings. Moreover,existing benchmarks often overlook a critical component of human ToM: behavioral prediction, the ability to use inferred mental states to guide strategic decision-making and select appropriate conversational actions for future interactions. To better align LLM-based ToM evaluation with human-like social reasoning, we propose RecToM, a novel benchmark for evaluating ToM abilities in recommendation dialogues. RecToM focuses on two complementary dimensions: Cognitive Inference and Behavioral Prediction. The former focus on understanding what has been communicated by inferring the underlying mental states. The latter emphasizes what should be done next, evaluating whether LLMs can leverage these inferred mental states to predict, select, and assess appropriate dialogue strategies. Together, these dimensions enable a comprehensive assessment of ToM reasoning in CRS. Extensive experiments on state-of-the-art LLMs demonstrate that RecToM poses a significant challenge. While the models exhibit partial competence in recognizing mental states, they struggle to maintain coherent, strategic ToM reasoning throughout dynamic recommendation dialogues, particularly in tracking evolving intentions and aligning conversational strategies with inferred mental states.

Longhao Li, Zhao Guo, Hongjie Chen, Yuhang Dai, Ziyu Zhang, Hongfei Xue, Tianlun Zuo, Chengyou Wang, Shuiyuan Wang, Xin Xu 等

The development of speech understanding and generation has been significantly accelerated by the availability of large-scale, high-quality speech datasets. Among these, ASR and TTS are regarded as the most established and fundamental tasks. However, for Cantonese (Yue Chinese), spoken by approximately 84.9 million native speakers worldwide, limited annotated resources have hindered progress and resulted in suboptimal ASR and TTS performance. To address this challenge, we propose WenetSpeech-Pipe, an integrated pipeline for building large-scale speech corpus with multi-dimensional annotation tailored for speech understanding and generation. Based on this pipeline, we release WenetSpeech-Yue, the first large-scale Cantonese speech corpus with multi-dimensional annotation for ASR and TTS, covering 21,800 hours across 10 domains with annotations including ASR transcription, text confidence, speaker identity, age, gender, speech quality scores, among other annotations. We also release WSYue-eval, a comprehensive Cantonese benchmark with two components: WSYue-ASR-eval, a manually annotated set for evaluating ASR on short and long utterances, code-switching, and diverse acoustic conditions, and WSYue-TTS-eval, with base and coverage subsets for standard and generalization testing. Experimental results show that models trained on WenetSpeech-Yue achieve competitive results against state-of-the-art (SOTA) Cantonese ASR and TTS systems, including commercial and LLM-based models, highlighting the value of our dataset and pipeline.

Junsong Li, Jie Zhou, Bihao Zhan, Yutao Yang, Qianjun Pan, Shilian Chen, Tianyu Huai, Xin Li, Qin Chen, Liang He

Alignment plays a crucial role in Large Language Models (LLMs) in aligning with human preferences on a specific task/domain. Traditional alignment methods suffer from catastrophic forgetting, where models lose previously learned values when adapting to new preferences or domains. We introduce LifeAlign, a novel framework for lifelong alignment that enables LLMs to maintain consistent human preference alignment across sequential learning tasks without forgetting previously learned values. Our approach consists of two key innovations. First, we propose a focalized preference optimization strategy that aligns LLMs with new preferences while preventing the erosion of alignment acquired from previous tasks. Second, we develop a short-to-long memory consolidation mechanism that merges denoised short-term preference representations into stable long-term memory using intrinsic dimensionality reduction, enabling efficient storage and retrieval of alignment patterns across diverse domains. We evaluate LifeAlign across multiple sequential alignment tasks spanning different domains and preference types. Experimental results demonstrate that our method achieves superior performance in maintaining both preference alignment quality and knowledge retention compared to existing lifelong learning approaches.