Papers by Xin Shen

ReFSQL: A Retrieval-Augmentation Framework for Text-to-SQL Generation (2023.findings-emnlp)

Challenge:	Existing methods to analyze black-box jailbreaks lack direct optimization signals to refine adversarial prompts.
Approach:	They propose a distribution-jailbreak attack method that selects effective jailbreak templates and iteratively optimizes adversarial suffixes by maximizing the KL divergence from the standard refusal distribution.
Outcome:	The proposed method achieves state-of-the-art Attack Success Rate (ASR) on all tested open-source models and delivers over 94% ASR on GPT-4.1.

Copied to clipboard

Kun Zhang, Xiexiong Lin, Yuanzhuo Wang, Xin Zhang, Fei Sun, Cen Jianhe, Hexiang Tan, Xuhui Jiang, Huawei Shen

Programming Languages in NLP

Challenge:	Existing methods that align natural language with SQL Language underestimate inherent structural characteristics of SQL and lead to structure errors.
Approach:	They propose a retrieval-argument framework that aligns natural language with SQL Language and trains one encoder-decoder-based model to fit all questions.
Outcome:	The proposed framework improves accuracy and robustness of text-to-SQL generation on five datasets.

BlindGuard: Safeguarding LLM-based Multi-Agent Systems under Unknown Attacks (2026.acl-long)

Copied to clipboard

Rui Miao, Yixin Liu, Yili Wang, Xu Shen, Yue Tan, Yiwei Dai, Shirui Pan, Xin Wang

Robustness in NLP

Challenge:	Existing supervised defense methods rely on labeled malicious agents to train a supervised model of malicious behavior.
Approach:	They propose an unsupervised defense method that learns without requiring any attack-specific labels or prior knowledge of malicious behaviors.
Outcome:	The proposed method detects diverse attack types across MAS with various communication patterns while maintaining superior generalizability compared to baselines.

Aligning VLM Assistants with Personalized Situated Cognition (2025.acl-long)

Copied to clipboard

Yongqi Li, Shen Zhou, Xiaohu Li, Xin Miao, Jintao Wen, Mayi Xu, Jianhao Chen, Birong Pan, Hankun Kang, Yuanyuan Zhu, Ming Zhong, Tieyun Qian

E-ConvRec: A Large-Scale Conversational Recommendation Dataset for E-Commerce Customer Service (2022.lrec-1)

Challenge:	Existing studies on vision-language models aligned with general human objectives have not been successful because people with diversified backgrounds have different cognition even in the same situation.
Approach:	They propose to characterize individuals based on the sociological concept of Role-Set and then evaluate their actions to see whether personalized alignment is achieved.
Outcome:	The proposed framework constructs a cognition-aware and action-based reward model for personalized alignment.

Copied to clipboard

Meihuizi Jia, Ruixue Liu, Peiying Wang, Yang Song, Zexi Xi, Haobin Li, Xin Shen, Meng Chen, Jinhui Pang, Xiaodong He

Natural Language Interfaces

Challenge:	Recent research has focused on developing conversational recommendation system (CRS), which provides valuable recommendations to users through conversations.
Approach:	They construct an authentic Chinese dialogue dataset consisting of over 25k dialogues and 770k utterances, which contains user profile, product knowledge base, and multiple sequential real conversations between users and recommenders.
Outcome:	The proposed dataset contains user profile, product knowledge base, and multiple sequential real conversations between users and recommenders.

Scientific Paper Extractive Summarization Enhanced by Citation Graphs (2022.emnlp-main)

Copied to clipboard

Xiuying Chen, Mingzhe Li, Shen Gao, Rui Yan, Xin Gao, Xiangliang Zhang

Information Extraction & Text Mining

Challenge:	citation graphs can be used to extract scientific papers under different conditions.
Approach:	They propose a multi-granularity unsupervised summarization model that fine tunes a pre-trained encoder model on the citation graph by link prediction tasks.
Outcome:	The proposed model outperforms baseline models on a public benchmark dataset.

Reasoning-Enhanced Domain-Adaptive Pretraining of Multimodal Large Language Models for Short Video Content Governance (2025.emnlp-industry)

Copied to clipboard

Zixuan Wang, Yu Sun, Hongwei Wang, Baoyu Jing, Xiang Shen, Xin Dong, Zhuolin Hao, Hongyu Xiong, Yang Song

An Ensemble-of-Experts Framework for Rehearsal-free Continual Relation Extraction (2024.findings-acl)

Challenge:	Existing approaches to identifying inappropriate content require extensive human-labeled data and lack cross-issue generalization.
Approach:	They propose a reasoning-enhanced multimodal large language model (MLLM) pretraining paradigm for unified inappropriate content detection.
Outcome:	The proposed model improves the MLLM's performance in both zero-shot and supervised fine-tuning settings and shows strong generalization capabilities to emergent, previously unseen issues.

Copied to clipboard

Shen Zhou, Yongqi Li, Xin Miao, Tieyun Qian

Information Extraction & Text Mining

Challenge:	Existing methods for continual relation extraction (CRE) are rehearsal-based and need to store samples and thus may encounter privacy and security issues.
Approach:	They propose an Ensemble-of-Experts framework for rehearsal-free continual relation extraction that discriminates between experts and augments analogous relations across tasks.
Outcome:	The proposed method outperforms existing rehearsal-free methods and is even better than existing methods.

Episodic Memory Retrieval from LLMs: A Neuromorphic Mechanism to Generate Commonsense Counterfactuals for Relation Extraction (2024.findings-acl)

Copied to clipboard

Xin Miao, Yongqi Li, Shen Zhou, Tieyun Qian

Information Extraction & Text Mining

Challenge:	Large language models (LLMs) have achieved satisfactory performance in counterfactual generation, however, there are misalignments between LLMs and humans which hinder LLM from handling complex tasks like relation extraction.
Approach:	They propose to mimic the episodic memory retrieval mechanism of human hippocampus to align LLMs’ generation process with that of humans.
Outcome:	The proposed framework improves over existing methods in terms of quality of counterfactuals.

SalaMAnder: Shapley-based Mathematical Expression Attribution and Metric for Chain-of-Thought Reasoning (2025.findings-emnlp)

Copied to clipboard

Yue Xin, Chen Shen, Shaotian Yan, Xiaosong Yuan, Yaoming Wang, Xiaofeng Zhang, Chenxi Huang, Jieping Ye

CharacterEval: A Chinese Benchmark for Role-Playing Conversational Agent Evaluation (2024.acl-long)

Challenge:	Chain-of-Thought prompting improves the math reasoning capability of large language models.
Approach:	They propose a method for attribution of component-level contributions in CoT reasoning using Shapley value and a stratified sampling algorithm that significantly reduces computational complexity.
Outcome:	The proposed method reduces computational complexity and provides robust correlations with model performance.

Copied to clipboard

Quan Tu, Shilong Fan, Zihang Tian, Tianhao Shen, Shuo Shang, Xin Gao, Rui Yan

Dialogue Systems & Conversational Agents

Challenge:	CharacterEval is a benchmark for comprehensive RPCA assessment in Chinese . authors show that Chinese LLMs exhibit more promising capabilities than GPT-4 in role-playing conversation.
Approach:	They propose a Chinese benchmark for comprehensive RPCA assessment . they use a dataset of Chinese role-playing dialogues and character profiles .
Outcome:	The proposed benchmark demonstrates that Chinese LLMs exhibit more promising capabilities than GPT-4 in Chinese role-playing conversation.

Evidence-Augmented Policy Optimization with Reward Co-Evolution for Long-Context Reasoning (2026.acl-long)

Copied to clipboard

Xin Guan, Zijian Li, Shen Huang, Pengjun Xie, Jingren Zhou, Jiuxin Cao

Reasoning

Challenge:	Evidence-Augmented Policy Optimization (EAPO) improves long-context reasoning performance . Xu et al., 2025): large language models are a critical part of NLP .
Approach:	They propose an Evidence-Augmented Reasoning paradigm that uses a group-relative reward to improve evidence quality.
Outcome:	EAPO significantly improves long-context reasoning performance compared to baselines.

Dialogue Summarization with Static-Dynamic Structure Fusion Graph (2023.acl-long)

Copied to clipboard

Shen Gao, Xin Cheng, Mingzhe Li, Xiuying Chen, Jinpeng Li, Dongyan Zhao, Rui Yan

Is the Attention Matrix Really the Key to Self-Attention in Multivariate Long-Term Time Series Forecasting? (2026.acl-long)

Challenge:	Dialogue summarization is a challenging task since it has dynamic interaction nature and inconsistent information flow among various speakers.
Approach:	They propose a Static-Dynamic graph-based Dialogue Summarization model which fuses prior knowledge from human expertise and adaptively learns the graph structure in an end-to-end learning fashion.
Outcome:	The proposed model can help people capture the highlights of a semi-structured and multi-participant dialogue without reviewing the complex dialogue context.

Copied to clipboard

Xinyu Li, Kexi Chen, Jiajie Shen, Ying Zheng, Hong Lu, Jin Zhao, Xin Wang

Prompting Large Language Models for Counterfactual Generation: An Empirical Study (2024.lrec-main)

Challenge:	In multivariate long-term time series forecasting, it is widely believed that the effectiveness of self-attention arises from its attention matrix.
Approach:	They propose a multi-branch MLP that isolates the ‘multi-brain mapping with element-wise operation’ structure from the Transformer and shows that it achieves competitive performance.
Outcome:	The proposed model outperforms three classic and three latest Transformer models and shows that it achieves competitive performance.

Copied to clipboard

Yongqi Li, Mayi Xu, Xin Miao, Shen Zhou, Tieyun Qian

Generate & Rank: A Multi-task Framework for Math Word Problems (2021.findings-emnlp)

Challenge:	Large language models (LLMs) have made remarkable progress in a wide range of natural language understanding and generation tasks, but their ability to generate counterfactuals has not been examined systematically.
Approach:	They propose a framework to evaluate LLMs' ability to generate counterfactuals based on key factors including intrinsic properties and prompt design.
Outcome:	The proposed framework examines the strengths and weaknesses of large language models (LLMs) and identifies factors that influence their ability to generate counterfactuals.

Copied to clipboard

Jianhao Shen, Yichun Yin, Lin Li, Lifeng Shang, Xin Jiang, Ming Zhang, Qun Liu

IPS: In-Prompt Process Supervision for Short Video Content Moderation (2026.acl-industry)

Challenge:	Existing studies formalize MWP as a generation task but mathematical expressions are prone to minor mistakes.
Approach:	They propose a ranking task for math word problem (MWP) that learns from its own mistakes and distinguishes between correct and incorrect expressions.
Outcome:	The proposed model outperforms baselines on the classical Math23k dataset and is 7% higher than the state-of-the-art.

Copied to clipboard

Mingchao Liu, Yu Sun, Ruixiao Sun, Xin Dong, Xiang Shen, Hongwei Wang, Hongyu Xiong, Yang Song

Too Consistent to Detect: A Study of Self-Consistent Errors in LLMs (2025.emnlp-main)

Challenge:	Multimodal large language models (MLLMs) capture semantics of short video content but fail to account for policy-specific details.
Approach:	They propose a framework that integrates In-prompt Process Supervision into MLLMs . they propose sequential reasoning over ancillary questions during fine-tuning .
Outcome:	IPS outperforms baseline MLLMs on public and proprietary benchmarks . replacing human-annotated ancillary labels with MLML-generated ones results in performance degradation.

Copied to clipboard

Hexiang Tan, Fei Sun, Sha Liu, Du Su, Qi Cao, Xin Chen, Jingang Wang, Xunliang Cai, Yuanzhuo Wang, Huawei Shen, Xueqi Cheng

Challenge:	Existing detection methods fail to account for self-consistent error . study identifies self-consistency errors and evaluates them .
Approach:	They propose a method that fuses hidden state evidence from an external verifier LLM to detect self-consistent errors.
Outcome:	The proposed method significantly enhances performance on self-consistent errors across three LLM families.

Multimodal Language Models See Better When They Look Shallower (2025.emnlp-main)

Copied to clipboard

Haoran Chen, Junyan Lin, Xinghao Chen, Yue Fan, Jianfeng Dong, Xin Jin, Hui Su, Jinlan Fu, Xiaoyu Shen

MEIT: Multimodal Electrocardiogram Instruction Tuning on Large Language Models for Report Generation (2025.findings-acl)

Challenge:	Existing studies show that multimodal large language models extract visual features from the final layers of a pretrained Vision Transformer.
Approach:	They propose a feature fusion method that strategically incorporates shallower layers . they propose MLLMs that extract visual features from the final layers of a pretrained Vision Transformer .
Outcome:	The proposed method outperforms deep layers on fine-grained visual tasks . it is the first comprehensive study of visual layer selection for MLLMs .

Copied to clipboard

Zhongwei Wan, Che Liu, Xin Wang, Chaofan Tao, Hui Shen, Jing Xiong, Rossella Arcucci, Huaxiu Yao, Mi Zhang

Truth or Sophistry? LoFa: A Benchmark for LLM Robustness Against Logical Fallacies (2026.acl-long)

Challenge:	Recent studies have focused on classifying cardiac conditions using ECG data but have overlooked ECG report generation, which is time-consuming and requires clinical expertise.
Approach:	They propose a Multimodal ECG Instruction Tuning framework that extends the capability of large language models (LLMs) for the task.
Outcome:	The proposed framework outperforms open-source LLMs and LLM backbones across two large-scale ECG datasets.

Copied to clipboard

Xudong Shen, li Yuan, Ye Chen, Xin Wu, Yi Cai, Zhiyong Wu

Robustness in NLP

Challenge:	Prior work has focused on the ability of Large Language Models to identify or classify fallacies, but their robustness against these fallacias in persuasive contexts remains largely unexplored.
Approach:	They propose a new metric to assess LLM robustness against fallacies by pairing factual questions with fallacious arguments and developing a multi-round debate framework to assess model resilience.
Outcome:	The proposed metric disentangles robustness from a model’s knowledge limitations and demonstrates unique vulnerability profiles across models.

BFS-Prover: Scalable Best-First Tree Search for LLM-based Automatic Theorem Proving (2025.acl-long)

Copied to clipboard

Ran Xin, Chenguang Xi, Jie Yang, Feng Chen, Hang Wu, Xia Xiao, Yifan Sun, Shen Zheng, Ming Ding

Information Retrieval

Challenge:	Existing approaches to theorem proving in large language models rely on value functions and/or Monte Carlo Tree Search (MCTS), but the potential of simpler methods like Best-First Tree Search remains underexplored.
Approach:	They propose a scalable expert iteration framework that implements strategic data filtering at each expert iteration round, excluding problems solvable via beam search node expansion to focus on harder cases.
Outcome:	The proposed framework achieves a state-of-the-art score of 72.95 on the MiniF2F test set and challenges the perceived necessity of complex tree search methods.

MEDA: Dynamic KV Cache Allocation for Efficient Multimodal Long-Context Inference (2025.naacl-long)

Copied to clipboard

Zhongwei Wan, Hui Shen, Xin Wang, Che Liu, Zheda Mai, Mi Zhang

PEGRL: Improving Machine Translation by Post-Editing Guided Reinforcement Learning (2026.findings-acl)

Challenge:	Long-context Multimodal Large Language Models (MLLMs) require substantial computational resources as their multimodal Key-Value (KV) cache grows with increasing input lengths, challenging memory and time efficiency.
Approach:	They propose a dynamic multimodal KV cache allocation strategy that dynamically allocating KV size based on attention entropy to better adapt to multimodal interactions.
Outcome:	The proposed model achieves up to 72% KV cache memory reduction and 2.82 faster decoding speeds while maintaining or enhancing performance on various multimodal tasks in a long context.

Copied to clipboard

Yunzhi Shen, Hao Zhou, Xin Huang, Xue Han, Junlan Feng, Shujian Huang

GiFT: Gibbs Fine-Tuning for Code Generation (2025.acl-long)

Challenge:	Reinforcement learning (RL) has shown strong promise for LLM-based machine translation . however, translation-oriented RL remains challenged by high-variance policy gradients induced by Monte Carlo baselines and large trajectory space that favors global exploration over fine-grained local optimization.
Approach:	They propose a two-stage RL framework that uses post-editing as an auxiliary task to stabilize training and guide overall optimization.
Outcome:	The proposed framework supports global exploration and fine-grained optimization while supporting global exploration.

Copied to clipboard

Haochen Li, Wanjin Feng, Xin Zhou, Zhiqi Shen

Programming Languages in NLP

Challenge:	Training Large Language Models (LLMs) with synthetic data is a prevalent practice in code generation.
Approach:	They propose a method to fine-tune large language models with code drawn from a conditional distribution, conditioned on a specific seed description.
Outcome:	The proposed method improves performance on four datasets and shows that it can be used to fine-tune LLMs with code derived from the marginal distribution.

VisFinEval: A Scenario-Driven Chinese Multimodal Benchmark for Holistic Financial Understanding (2025.emnlp-main)

Copied to clipboard

Zhaowei Liu, Xin Guo, Haotian Xia, Lingfeng Zeng, Fangqi Lou, Jinyi Niu, Mengping Li, Qi Qi, Jiahuan Li, Wei Zhang, Yinglong Wang, Weige Cai, Weining Shen, Liwen Zhang

Enhancing Chain-of-Thought Reasoning with Critical Representation Fine-tuning (2025.acl-long)

Challenge:	Existing benchmarks focus on text comprehension, but MLLMs lack the ability to integrate visual data over financial visuals.
Approach:	They evaluate 21 state-of-the-art multimodal large language models in a zero-shot setting . they use an annotated question–answer pair from eight common financial image modalities .
Outcome:	The new benchmark outperforms existing models but trailed financial experts by 14 percentage points.

Copied to clipboard

Chenxi Huang, Shaotian Yan, Liang Xie, Binbin Lin, Sinan Fan, Yue Xin, Deng Cai, Chen Shen, Jieping Ye

Semantic Text Processing

Challenge:	Representation Fine-tuning (ReFT) is a proposed method for improving parameter efficiency . however, it yields suboptimal performance, as fixed-position representations have uncertain impact on outputs .
Approach:	They propose a method that fine-tunes critical representations in a low-rank linear subspace while freezing the base model.
Outcome:	The proposed method improves accuracy of LLaMA-2-7B and ReFT by 18.2 and 3.8 on GSM8K.

SeaLLMs - Large Language Models for Southeast Asia (2024.acl-demos)

Copied to clipboard

Xuan-Phi Nguyen, Wenxuan Zhang, Xin Li, Mahani Aljunied, Zhiqiang Hu, Chenhui Shen, Yew Ken Chia, Xingxuan Li, Jianyu Wang, Qingyu Tan, Liying Cheng, Guanzheng Chen, Yue Deng, Sen Yang, Chaoqun Liu, Hang Zhang, Lidong Bing

Rewriting the Code: A Simple Method for Large Language Model Augmented Code Search (2024.acl-long)

Challenge:	Existing large language models favor high-resource languages, such as English, at the expense of low-resourced and regional languages.
Approach:	They propose a series of language models that specifically focuses on Southeast Asian languages.
Outcome:	SeaLLM models outperform ChatGPT-3.5 in non-Latin languages by large margins . linguistic disparity impedes access to state-of-the-art AI technologies for non-English-speaking populations .

Copied to clipboard

Haochen Li, Xin Zhou, Zhiqi Shen

Semantic Text Processing

Challenge:	Experimental results show that ReCo significantly boosts retrieval accuracy across sparse, zero-shot dense and fine-tuned dense search settings.
Approach:	They propose a generation-augmented retrieval framework that additionally Rewrites the Code (ReCo) within the codebase for style normalization.
Outcome:	The proposed method significantly boosts retrieval accuracy across sparse, zero-shot dense, and fine-tuned dense retrieval settings in diverse search scenarios.

FinEval: A Chinese Financial Domain Knowledge Evaluation Benchmark for Large Language Models (2025.naacl-long)

Copied to clipboard

Xin Guo, Haotian Xia, Zhaowei Liu, Hanyang Cao, Zhi Yang, Zhiqiang Liu, Sizhe Wang, Jinyi Niu, Chuqi Wang, Yanhui Wang, Xiaolong Liang, Xiaoming Huang, Bing Zhu, Zhongyu Wei, Yun Chen, Weining Shen, Liwen Zhang

SVD-LLM V2: Optimizing Singular Value Truncation for Large Language Model Compression (2025.naacl-long)

Challenge:	Large language models have demonstrated outstanding performance in various natural language processing tasks, but their security capabilities in the financial domain have not been explored.
Approach:	They propose to use a benchmark to evaluate large language models' financial domain knowledge and practical abilities.
Outcome:	The proposed benchmark evaluates large language models' financial domain knowledge and practical abilities.

Copied to clipboard

Xin Wang, Samiul Alam, Zhongwei Wan, Hui Shen, Mi Zhang

Embodied-Reasoner: Synergizing Visual Search, Reasoning, and Action for Embodied Interactive Tasks (2026.acl-long)

Challenge:	Existing methods for compressing Large Language Models suffer from significant truncation losses.
Approach:	They propose a novel method that optimizes singular value truncation in SVD compression . they use dynamic compression ratio allocation to balance the large tuncation loss .
Outcome:	The proposed method outperforms current state-of-the-art methods on ten datasets and five models on various scales.

Copied to clipboard

Wenqi Zhang, Mengna Wang, Gangao Liu, Huixin Xu, Yiwei Jiang, Yongliang Shen, Guiyang Hou, Zhe Zheng, Hang Zhang, Xin Li, Jiajun Liu, Weiming Lu, Peng Li, Yueting Zhuang

Reasoning

Challenge:	Recent advances in reasoning models have demonstrated remarkable capabilities on mathematical and coding tasks, but their effectiveness in embodied domains remains largely unexplored.
Approach:	They propose a reasoning model for interactive embodied tasks that synthesizes 9.3k coherent Observation-Thought-Action trajectories containing 64k ego-centric images and 90k diverse reasoning processes.
Outcome:	The proposed model outperforms existing visual reasoning models by +9%, 24%, and +13% on long-horizon tasks.

Whether LLMs Know If They Know: Identifying Knowledge Boundaries via Debiased Historical In-Context Learning (2025.findings-acl)

Copied to clipboard

Bo Lv, Nayu Liu, Yang Shen, Xin Liu, Ping Luo, Yue Yu

Towards Generating Long and Coherent Text with Multi-Level Latent Variable Models (P19-1)

Challenge:	Existing methods for active retrieval (AR) rely on training classification models or using the confidence of the model’s answer to determine knowledge boundaries.
Approach:	They propose a method to identify knowledge boundaries in active retrieval by retrieving historical queries as high-confidence in-context examples.
Outcome:	Experiments on four QA benchmarks show that DH-ICL achieves performance comparable to full retrieval on LLaMA with only half the number of retrievals, without any additional training.

Copied to clipboard

Dinghan Shen, Asli Celikyilmaz, Yizhe Zhang, Liqun Chen, Xin Wang, Jianfeng Gao, Lawrence Carin

Optimize Weight Rounding via Signed Gradient Descent for the Quantization of LLMs (2024.findings-emnlp)

Challenge:	Variational autoencoders (VAEs) have received much attention as an end-to-end architecture for text generation with latent variables.
Approach:	They propose to leverage several multi-level structures to learn a variational autoencoder model for generating long, and coherent text.
Outcome:	The proposed model produces more coherent and less repetitive long text compared to baselines and mitigates posterior collapse issue.

Copied to clipboard

Wenhua Cheng, Weiwei Zhang, Haihao Shen, Yiyang Cai, Xin He, Lv Kaokao, Yi Liu