Papers by Huang Fang

Thinking with DistilQwen: A Tale of Four Distilled Reasoning and Reward Model Series (2025.emnlp-industry)

Challenge:	Existing methods focus on replicating dialogues in textual form, neglecting the role’s voice traits as a crucial effect in interaction, which tends to be more immersive experiences in realistic scenarios.
Approach:	They propose a first seamless speech-language personality interaction model to achieve immersive RPAs with low latency.
Outcome:	The proposed model exhibits role-specific personality traits and vocal traits throughout the interaction, enabling a mixture of speech and language responses.

Copied to clipboard

Wenrui Cai, Chengyu Wang, Junbing Yan, Jun Huang, Xiangzhong Fang

Green & Sustainable NLP

Challenge:	In the rapidly evolving landscape of large language models, the need for efficient reasoning models has become increasingly urgent.
Approach:	They extend the Qwen model family by introducing four model series specifically designed for industrial applications.
Outcome:	The proposed models outperform previous models in multiple benchmarks and provide scalable training and inference functionality on the Alibaba Cloud PAI platform.

CART: A Generative Cross-Modal Retrieval Framework With Coarse-To-Fine Semantic Modeling (2025.acl-long)

Copied to clipboard

Minghui Fang, Shengpeng Ji, Jialong Zuo, Hai Huang, Yan Xia, Jieming Zhu, Xize Cheng, Xiaoda Yang, Wenrui Liu, Gang Wang, Zhenhua Dong, Zhou Zhao

Model Composition for Multimodal Large Language Models (2024.acl-long)

Challenge:	Cross-modal retrieval tasks are used to retrieve data from one modality or another based on a query from another modality.
Approach:	They propose a generative cross-modal retrieval framework based on coarse-to-fine semantic modeling . they propose combining K-Means and RQ-VAE to discretize multimodal data into token sequences that support autoregressive generation.
Outcome:	The proposed framework achieves excellent performance and efficiency in multimodal retrieval tasks.

Copied to clipboard

Chi Chen, Yiyang Du, Zheng Fang, Ziyue Wang, Fuwen Luo, Peng Li, Ming Yan, Ji Zhang, Fei Huang, Maosong Sun, Yang Liu

LLMs Assist NLP Researchers: Critique Paper (Meta-)Reviewing (2024.emnlp-main)

Challenge:	Existing methods for creating versatile MLLMs rely on joint training with paired instruction data, which is resource-intensive and challenging to extend to new modalities.
Approach:	They propose a new paradigm for multimodal large language models by reusing modality encoders and merging LLM parameters.
Outcome:	The proposed model retains the modal understanding capabilities of each original model.

Copied to clipboard

Jiangshu Du, Yibo Wang, Wenting Zhao, Zhongfen Deng, Shuaiqi Liu, Renze Lou, Henry Zou, Pranav Narayanan Venkit, Nan Zhang, Mukund Srinath, Haoran Zhang, Vipul Gupta, Yinghui Li, Tao Li, Fei Wang, Qin Liu, Tianlin Liu, Pengzhi Gao, Congying Xia, Chen Xing, Cheng Jiayang, Zhaowei Wang, Ying Su, Raj Shah, Ruohao Guo, Jing Gu, Haoran Li, Kangda Wei, Zihao Wang, Lu Cheng, Surangika Ranathunga, Meng Fang, Jie Fu, Fei Liu, Ruihong Huang, Eduardo Blanco, Yixin Cao, Rui Zhang, Philip Yu, Wenpeng Yin

Challenge:	a comparative analysis of paper (meta-)reviews by large language models (LLMs) aims to identify and distinguish LLMs from human activities .
Approach:	They present a comparative analysis to identify and distinguish LLM activities from human activities.
Outcome:	The proposed analysis aims to improve recognition of instances when someone implicitly uses LLMs for reviewing activities.

CRITICTOOL: Evaluating Self-Critique Capabilities of Large Language Models in Tool-Calling Error Scenarios (2025.emnlp-main)

Copied to clipboard

Shiting Huang, Zhen Fang, Zehui Chen, Siyu Yuan, Junjie Ye, Yu Zeng, Lin Chen, Qi Mao, Feng Zhao

TrInk: Ink Generation with Transformer Network (2025.emnlp-main)

Challenge:	a number of tools are used to perform complex tasks, but the tool utilization process can cause errors.
Approach:	They propose a critique evaluation benchmark for tool learning that analyzes function-calling errors on tool evaluation benchmarks.
Outcome:	The proposed critique evaluation benchmark holds diverse tool-use errors with varying complexities, which better reflects real-world scenarios.

Copied to clipboard

Zezhong Jin, Shubhang Desai, Xu Chen, Biyi Fang, Zhuoyi Huang, Zhe Li, Chong-Xin Gan, Xiao Tu, Man-Wai Mak, Yan Lu, Shujie Liu

Context-DPO: Aligning Language Models for Context-Faithfulness (2025.findings-acl)

Challenge:	Existing methods for handwriting generation capture global dependencies and can generate high-quality handwritten samples.
Approach:	They propose a Transformer-based model for ink generation, TrInk, which captures global dependencies.
Outcome:	The proposed model reduces character error rate and word error rate by 35.56% on the IAM-OnDB dataset compared to previous models.

Copied to clipboard

Baolong Bi, Shaohan Huang, Yiwei Wang, Tianchi Yang, Zihan Zhang, Haizhen Huang, Lingrui Mei, Junfeng Fang, Zehao Li, Furu Wei, Weiwei Deng, Feng Sun, Qi Zhang, Shenghua Liu

Language-Codec: Bridging Discrete Codec Representations and Speech Language Models (2025.acl-long)

Challenge:	Context-DPO is the first alignment method specifically designed to enhance contextfaithfulness for large language models.
Approach:	They propose a benchmark that simulates Retrieval-Augmented Generation scenarios with knowledge conflicts to evaluate context-faithfulness.
Outcome:	The proposed method improves LLMs' context-faithfulness by 35% to 280% over open-source models.

Copied to clipboard

Shengpeng Ji, Minghui Fang, Jialong Zuo, Ziyue Jiang, Dingdong Wang, Hanting Wang, Hai Huang, Zhou Zhao

DevEval: A Manually-Annotated Code Generation Benchmark Aligned with Real-World Code Repositories (2024.findings-acl)

Challenge:	Existing gaps between discrete acoustic codecs and downstream speech language models . initial channel of codebooks contains excessive information, making it difficult to generate tokens from weakly supervised signals such as text.
Approach:	They propose a discrete acoustic codec for generating acustic tokens from weakly supervised signals.
Outcome:	The proposed language-codec outperforms competing audio compression algorithms and validates on downstream speech language models.

Copied to clipboard

Jia Li, Ge Li, Yunfei Zhao, Yongmin Li, Huanyu Liu, Hao Zhu, Lecheng Wang, Kaibo Liu, Zheng Fang, Lanshen Wang, Jiazheng Ding, Xuanming Zhang, Yuqi Zhu, Yihong Dong, Zhi Jin, Binhua Li, Fei Huang, Yongbin Li, Bin Gu, Mengfei Yang

Weakly-Supervised Aspect-Based Sentiment Analysis via Joint Aspect-Sentiment Topic Embedding (2020.emnlp-main)

Challenge:	Existing benchmarks are poorly aligned with real-world code repositories and are insufficient to evaluate the coding abilities of Large Language Models (LLMs).
Approach:	They propose a repository-level benchmark named DevEval to evaluate LLMs' coding abilities in real-world code repositories.
Outcome:	The proposed benchmarks show that the LLMs perform better in real-world code repositories than existing benchmarks.

Copied to clipboard

Jiaxin Huang, Yu Meng, Fang Guo, Heng Ji, Jiawei Han

Spiral of Silence in Large Language Model Agents (2025.findings-emnlp)

Challenge:	Existing methods for aspect-based sentiment analysis of review text use only a few keywords describing each aspect/sentiment without using any labeled examples.
Approach:	They propose a weakly-supervised approach for aspect-based sentiment analysis which uses only a few keywords describing each aspect/sentiment without using any labeled examples.
Outcome:	The proposed method generates quality joint topics and outperforms baselines significantly on benchmark datasets.

Copied to clipboard

Mingze Zhong, Meng Fang, Zijing Shi, Yuxuan Huang, Shunfeng Zheng, Yali Du, Ling Chen, Jun Wang

C-World: A Computer Use Agent Environment Creator (2026.acl-long)

Challenge:	Existing theories of Spiral of Silence do not apply to large language models .
Approach:	They propose an evaluation framework for examining SoS in large language models . they consider four controlled conditions that vary the availability of "History" and "Persona" signals .
Outcome:	The proposed framework examines the SoS-like dynamics in large language models . it shows that history and persona together produce strong majority dominance .

Copied to clipboard

Ziqiao Xi, Shuang Liang, Qi Liu, Jiaqing Zhang, Letian Peng, Fang Nan, Meshal Nayim, Tianhui Zhang, Rishika Mundada, Lianhui Qin, Biwei Huang, Kun Zhou

Challenge:	C-World enables users to build agent environments on demand.
Approach:	They propose a system that enables users to build agent environments on demand.
Outcome:	The proposed system outperforms baselines on 119k samples and achieves Spearman = 0.883 ranking correlation with real execution.

If an LLM Were a Character, Would It Know Its Own Story? Evaluating Lifelong Learning in LLMs (2026.acl-long)

Copied to clipboard

Siqi Fan, Xiusheng Huang, Yiqun Yao, Xuezhi Fang, Kang Liu, Peng Han, Shuo Shang, Aixin Sun, Yequan Wang

Challenge:	Existing benchmarks for large language models (LLMs) fail to capture these dynamics, focusing on static, open-ended evaluations.
Approach:	They propose a benchmark to assess lifelong learning in large language models . they use two episodic datasets rich in narrative structure and character interactions .
Outcome:	Experiments on LLMs show that non-parametric methods outperform parametric ones in managing stateful learning.

Enhancing Reasoning Abilities of Small LLMs with Cognitive Alignment (2025.emnlp-main)

Copied to clipboard

Wenrui Cai, Chengyu Wang, Junbing Yan, Jun Huang, Xiangzhong Fang

Reasoning

Challenge:	Existing methods to distill chain-of-thought (CoT) results from large language reasoning models (LRMs) to small models are ineffective and require substantial amount of annotated data.
Approach:	They propose a Critique-Rethink-Verify system for training small language reasoning models that can be critiquized according to the cognitive capabilities of smaller models.
Outcome:	The proposed system outperforms other methods on challenging reasoning benchmarks.

A Mixed-Language Multi-Document News Summarization Dataset and a Graphs-Based Extract-Generate Model (2025.naacl-long)

Copied to clipboard

Shengxiang Gao, Fang Nan, Yongbing Zhang, Yuxin Huang, Kaiwen Tan, Zhengtao Yu

Information Extraction & Text Mining

Challenge:	Existing research on news summarization focuses on single-language single-document (SLSD), single-linguistic multi-document or cross-language multi-doc (CLSD) however, in real-world scenarios, news articles often involve multiple documents in different languages, i.e., mixed-language MLMD.
Approach:	They propose a mixed-language multi-document news summarization dataset with four different languages and 10,992 source document cluster and target summary pairs.
Outcome:	The proposed dataset contains four different languages and 10,992 source document cluster and target summary pairs.

Towards General Agentic Intelligence via Environment Scaling (2026.findings-acl)

Copied to clipboard

Runnan Fang, Shihao Cai, Baixuan Li, Jialong Wu, Guangyu Li, Wenbiao Yin, Xinyu Wang, Xiaobin Wang, Liangcai Su, Zhen Zhang, Shibin Wu, Zhengwei Tao, Yong Jiang, Pengjun Xie, Ningyu Zhang, Fei Huang, Wentao Zhang, Jingren Zhou

Infinity-Parser: Layout-Aware Reinforcement Learning with High-quality Document Parsing Dataset (2026.findings-acl)

Challenge:	Diverse real-world APIs require precise, robust function-calling intelligence, which needs agents to develop these capabilities through interaction in varied environments.
Approach:	They propose a framework that scales up environments to enable agentic intelligence . they use a two-phase agent fine-tuning strategy to first endow agents with basic agentic capabilities, then specializing them for domain-specific contexts.
Outcome:	Experiments on -bench, -Bench, and ACEBench show that the model significantly enhances the models’ function-calling capability.

Copied to clipboard

Baode Wang, Biao Wu, Weizhen Li, Meng Fang, Zuming Huang, Jun Huang, Yanjie Liang, Haozhe Wang, Ling Chen, Wei Chu, Yuan Qi

SciAssess: Benchmarking LLM Proficiency in Scientific Literature Analysis (2025.findings-naacl)

Challenge:	Existing supervised fine-tuning methods struggle to generalize across document types, leading to poor performance.
Approach:	They propose layoutRL, a reinforcement learning framework that optimizes layout understanding through composite rewards integrating normalized edit distance, paragraph count accuracy, and reading order preservation.
Outcome:	The proposed model outperforms specialized document parsing systems and general-purpose vision-language models on a broad range of document types, languages, and structural complexities.

Copied to clipboard

Hengxing Cai, Xiaochen Cai, Junhan Chang, Sihang Li, Lin Yao, Wang Changxin, Zhifeng Gao, Hongshuai Wang, Li Yongge, Mujie Lin, Shuwen Yang, Jiankun Wang, Mingjun Xu, Jin Huang, Xi Fang, Jiaxi Zhuang, Yuqi Yin, Yaqi Li, Changhong Chen, Zheng Cheng, Zifeng Zhao, Linfeng Zhang, Guolin Ke

Challenge:	Existing benchmarks fail to adequately evaluate the proficiency of Large Language Models (LLMs) Existing standards do not cover the skills needed to evaluate LLMs in scientific literature analysis.
Approach:	They propose a benchmark to evaluate the proficiency of large language models in scientific literature analysis.
Outcome:	SciAssess evaluates 11 LLMs on multiple tasks across scientific fields.

i-Code V2: An Autoregressive Generation Framework over Vision, Language, and Speech Data (2024.findings-naacl)

Copied to clipboard

Ziyi Yang, Mahmoud Khademi, Yichong Xu, Reid Pryzant, Yuwei Fang, Chenguang Zhu, Dongdong Chen, Yao Qian, Xuemei Gao, Yi-Ling Chen, Robert Gmyr, Naoyuki Kanda, Noel Codella, Bin Xiao, Yu Shi, Lu Yuan, Takuya Yoshioka, Michael Zeng, Xuedong Huang

LLM-Based Human-Agent Collaboration and Interaction Systems: A Survey (2026.findings-acl)

Challenge:	i-Code V2 is one of the first models capable of generating natural language from any combination of Vision, Language, and Speech data.
Approach:	They propose to create a model that can generate natural language from any combination of Vision, Language, and Speech data.
Outcome:	i-Code V2 matches or outperforms state-of-the-art single- and dual-modality baselines on 7 multimodal tasks.

Copied to clipboard

Henry Peng Zou, Wei-Chieh Huang, Yaozu Wu, Jizhou Guo, Yankai Chen, Chunyu Miao, Hoang H Nguyen, Yue Zhou, Weizhi Zhang, Liancheng Fang, Hanrong Zhang, Fangxin Wang, Pengfei Zhang, Langzhou He, Yangning Li, Dongyuan Li, Renhe Jiang, Philip S. Yu

Act-Adaptive Margin: Dynamically Calibrating Reward Models for Subjective Ambiguity (2026.acl-long)

Challenge:	Recent advances in large language models (LLMs) have sparked growing interest in building fully autonomous agents.
Approach:	They propose to integrate human-provided information, feedback, or control into the agent system to enhance system performance, reliability, and safety.
Outcome:	The proposed systems improve system performance, reliability, and safety by integrating human-provided information, feedback, or control into the agent system.

Copied to clipboard

Feiteng Fang, Dingwei Chen, Xiang Huang, Ting-En Lin, Yuchuan Wu, Xiong Liu, Jing Ye, Ziqiang Liu, Haonan Zhang, Liang Zhu, Hamid Alinejad-Rokny, Min Yang, Yongbin Li

Challenge:	Existing approaches to reward modeling in reinforcement learning tasks are limited when dealing with ambiguous preferences.
Approach:	They propose to use AAM to dynamically calibrate preference margins using the Bradley-Terry model's internal parameter knowledge to improve reward modeling in subjective tasks.
Outcome:	The proposed approach improves reward modeling by dynamically calibrating preference margins using the model’s internal parameter knowledge.

Memp: Exploring Agent Procedural Memory (2026.findings-acl)

Copied to clipboard

Runnan Fang, Yuan Liang, Xiaobin Wang, Jialong Wu, Shuofei Qiao, Pengjun Xie, Fei Huang, Huajun Chen, Ningyu Zhang

Challenge:	Large Language Models (LLMs) based agents suffer from brittle procedural memory that is manually engineered or entangled in static parameters.
Approach:	They propose a procedural-memory repository that distills past agent trajectories into fine-grained, step-by-step instructions and higher-level, script-like abstractions.
Outcome:	The proposed repository can be used to improve agents' performance on travelplanner and Alfworld.

DAGN: Discourse-Aware Graph Network for Logical Reasoning (2021.naacl-main)

Copied to clipboard

Yinya Huang, Meng Fang, Yu Cao, Liwei Wang, Xiaodan Liang

Structured Data in NLP

Challenge:	Recent QA with logical reasoning questions requires passage-level relations among the sentences.
Approach:	They propose a discourse-aware graph network that aggregates passage-level clues for QA by using discourse-based information.
Outcome:	The proposed model achieves competitive results on two logical reasoning QA datasets.

OmniThink: Expanding Knowledge Boundaries in Machine Writing through Thinking (2025.emnlp-main)

Copied to clipboard

Zekun Xi, Wenbiao Yin, Jizhan Fang, Jialong Wu, Runnan Fang, Yong Jiang, Pengjun Xie, Fei Huang, Huajun Chen, Ningyu Zhang

Information Retrieval

Challenge:	Recent advances in Large Language Models (LLMs) have demonstrated remarkable progress in machine writing such as open domain long-form generation.
Approach:	They propose a slow-thinking machine writing framework that emulates the human-like process of iterative expansion and reflection to improve the knowledge density of generated articles.
Outcome:	The proposed framework improves the knowledge density of generated articles without compromising metrics such as coherence and depth.

Beyond Quantity: Trajectory Diversity Scaling for Code Agents (2026.findings-acl)

Copied to clipboard

Guhong Chen, Chenghao Sun, Cheng Fu, Qiyao Wang, Zhihong Huang, ChaoPeng Wei, Guangxu Chen, Feiteng Fang, Ahmadreza Argha, Bing Zhao, Xander Xu, Qi Han, Hamid Alinejad-Rokny, Qiang Qu, Binhua Li, Shiwen Ni, Min Yang, HU Wei, Yongbin Li

Green & Sustainable NLP

Challenge:	Code large language models (LLMs) are becoming tool-interactive agents . quantity-centric scaling exhibits an early bottleneck that underutilizes trajectory data . et al.: a new approach to scale trajectory diversity improves tool-use generalization .
Approach:	They propose a Trajectory Diversity Scaling-based data synthesis framework for code agents that scales performance through diversity rather than raw volume.
Outcome:	Experiments on general tool-use benchmarks and code agent tasks show that TDScaling improves tool-user generalization and inherent coding proficiency.

Ouroboros: Generating Longer Drafts Phrase by Phrase for Faster Speculative Decoding (2024.emnlp-main)

Copied to clipboard

Weilin Zhao, Yuxiang Huang, Xu Han, Wang Xu, Chaojun Xiao, Xinrong Zhang, Yewei Fang, Kaihuo Zhang, Zhiyuan Liu, Maosong Sun

Challenge:	Speculative decoding is a widely used method that accelerates the generation process of large language models (LLMs) drafting efficiency has become a bottleneck in the final speedup of speculative drafting, therefore generating longer drafts at less cost can lead to better speedup.
Approach:	They propose a method that uses existing model to drafting and target LLM to verify draft in a low-cost parallel manner.
Outcome:	The proposed method can achieve speedups of up to 2.4 over speculative decoding and 3.9 over vanilla decoding without fine-tuning draft and target models.

Scalable-DSC: A Structural Template Prompt Approach to Scalable Dialogue State Correction (2023.emnlp-main)

Copied to clipboard

Haoxiang Su, Hongyan Xie, Hao Huang, Shuangyong Song, Ruiyu Fang, Xiaomeng Huang, Sijie Feng

Dialogue Systems & Conversational Agents

Challenge:	Existing approaches to correct wrong slot values in dialogue state tracking are intertwined with specific DST models, limiting their applicability to other DSTs.
Approach:	They propose a Scalable Dialogue State Correction model that corrects wrong slot values in predicted dialogue states by using a structural template prompt.
Outcome:	The proposed model achieves state-of-the-art results on MultiWOZ 2.0-2.4.

Reinforcement Learning on Pre-Training Data (2026.acl-long)

Copied to clipboard

Siheng Li, Kejiao Li, Zenan Xu, Guanhua Huang, Kun Li, Haoyuan Wu, null Wujiajia, Zihao Zheng, Chenchen Zhang, Kun Shi, Xue Gong, Qi Yi, Ruibin Xiong, Tingqiang Xu, Yuhao Jiang, Jianfeng Yan, Yuyuan Zeng, Guanghui Xu, Jinbao Xue, Zhijiang xu, Zheng Fang, Shuai LI, Qibin Liu, Xiaoxue Li, Zhuoyu Li, Yangyu Tao, Fei Gao, Cheng Jiang, Bochao Wang, Kai Liu, Jianchen Zhu, Wai Lam, Bo Zhou, Di Wang

Enhancing Personalized Dialogue Generation with Contrastive Latent Variables: Combining Sparse and Dense Persona (2023.acl-long)

Challenge:	Recent progress in large language models is driven by scaling of training compute through pre-training with nexttoken prediction (NTP) or post-training (RL) Pre-training using NTP enables models to acquire extensive knowledge and skills from general data, but it suffers from data inefficiency and catastrophic forgetting in continual learning settings.
Approach:	They propose to scale training compute through pre-training with next-token prediction (NTP) or post-training by scaling reinforcement learning (RL) to improve learning from general data.
Outcome:	Experiments on multiple benchmarks and models show that the proposed approach improves continual pre-training and provides a strong foundation for post-training on Qwen3-8B-Base.

Copied to clipboard

Yihong Tang, Bo Wang, Miao Fang, Dongming Zhao, Kun Huang, Ruifang He, Yuexian Hou

Dialogue Systems & Conversational Agents

Challenge:	Existing personalized dialogue agents model persona profiles from sparse or dense persona descriptions and dialogue histories.
Approach:	They propose a model that clusters dense persona descriptions into sparse categories and generates personalized responses from dialogue histories.
Outcome:	The proposed model improves on Chinese and English datasets.

Analyzing the Intensity of Complaints on Social Media (2022.findings-naacl)

Copied to clipboard

Ming Fang, Shi Zong, Jing Li, Xinyu Dai, Shujian Huang, Jiajun Chen

Cross-Lingual Transfer

Challenge:	Prior studies on identifying the existence or the type of complaints focus on building automatic classification models for identifying complaints.
Approach:	They propose to measure the intensity of complaints from text using Best-Worst Scaling method to estimate the popularity of posts on social media.
Outcome:	The proposed model can estimate the popularity of complaints on social media with best-worst scaling (BWS) method.

MathCanvas: Intrinsic Visual Chain-of-Thought for Multimodal Mathematical Reasoning (2026.acl-long)

Copied to clipboard

Weikang Shi, Aldrich Yu, Rongyao Fang, Houxing Ren, Ke Wang, Aojun Zhou, Changyao Tian, Xinyu Fu, Yuxuan Hu, Zimu Lu, Linjiang Huang, Si Liu, Rui Liu, Hongsheng Li

Beyond Accuracy: Unveiling Inefficiency Patterns in Tool-Integrated Reasoning (2026.acl-long)

Challenge:	Existing approaches to visual chain-of-thought are limited by external tools or fail to generate high-fidelity diagrams.
Approach:	They propose a framework to enable large multimodal models with VCoT capabilities . they pre-train a model on a 15.2M-pair corpus and teach it how to leverage visual aids .
Outcome:	The proposed framework unlocks complex, human-like visual reasoning in large language models . it pre-trains the model on a 15.2M-pair corpus and fine-tunes it on MathCanvas-Instruct .

Copied to clipboard

Qisheng Su, Shiting Huang, Zhen Fang, Ziyan Chen, Zehui Chen, Feng Zhao

Green & Sustainable NLP

Challenge:	Tool-Integrated Reasoning (TIR) is a tool that can be used to solve complex tasks.
Approach:	They propose a hardware-aware TIR-efficiency metric that unifies internal reasoning and external tool-use costs while explicitly accounting for non-reusable KV-Cache and long-tool-response scenarios.
Outcome:	The proposed metric explains wall-clock latency significantly better than token-count metric in a simulated high-concurrency industrial setting.

Editing the Mind of Giants: An In-Depth Exploration of Pitfalls of Knowledge Editing in Large Language Models (2024.findings-emnlp)

Copied to clipboard

Cheng-Hsun Hsueh, Paul Huang, Tzu-Han Lin, Che Liao, Hung-Chieh Fang, Chao-Wei Huang, Yun-Nung Chen

VaseVQA: Multimodal Agent and Benchmark for Ancient Greek Pottery (2026.findings-eacl)

Challenge:	Knowledge editing is a promising technique for updating factual knowledge in large language models (LLMs) but studies have identified side effects such as knowledge distortion and the deterioration of general abilities that have emerged after editing.
Approach:	They propose to evaluate the side effects of knowledge editing in large language models using metrics and benchmarks.
Outcome:	The results of the study highlight the limitations of current knowledge editing methods and outline potential research directions.

Copied to clipboard

Jinchao Ge, Tengfei Cheng, Biao Wu, Zeyu Zhang, Shiya Huang, Judith Bishop, Gillian Shepherd, Meng Fang, Ling Chen, Yang Zhao

NL2Lean: Translating Natural Language into Lean 4 through Multi-Aspect Reinforcement Learning (2025.emnlp-main)

Challenge:	MLLMs that use domain-specific data are limited in understanding cultural heritage artifacts such as ancient Greek pottery . supervised fine-tuning improves adaptation to domain knowledge, but it struggles with deeper reasoning tasks.
Approach:	They propose a visual question-answer tool that augments SFT with reinforcement learning using verifiable rewards.
Outcome:	The proposed model outperforms baseline models on reasoning-intensive questions on ancient Greek pottery.

Copied to clipboard

Yue Fang, Shaohan Huang, Xin Yu, Haizhen Huang, Zihan Zhang, Weiwei Deng, Furu Wei, Feng Sun, Qi Zhang, Zhi Jin

Machine Translation

Challenge:	Existing formal proof assistants rely on instruction tuning and lack fine-grained structural and semantic alignment.
Approach:	They propose a reinforcement learning framework that enables LLMs to translate natural language into formal language such as Lean 4 . they use a model with basic translation ability to refine the model's reinforcement learning .
Outcome:	The proposed method outperforms baseline models on NL-to-Lean 4 tasks.

Can MLLMs Understand the Deep Implication Behind Chinese Images? (2025.acl-long)

Copied to clipboard

Chenhao Zhang, Xi Feng, Yuelin Bai, Xeron Du, Jinchang Hou, Kaixin Deng, Guangzeng Han, Qinrui Li, Bingli Wang, Jiaheng Liu, Xingwei Qu, Yifei Zhang, Qixuan Zhao, Yiming Liang, Ziqiang Liu, Feiteng Fang, Min Yang, Wenhao Huang, Chenghua Lin, Ge Zhang, Shiwen Ni

WebWalker: Benchmarking LLMs in Web Traversal (2025.acl-long)

Challenge:	MLLMs perform poorly on traditional culture images, indicating limitations in understanding high-level semantics and lacking a deep knowledge base of Chinese traditional culture.
Approach:	They propose to use Chinese images to assess MLLMs' higher-order perception and understanding of Chinese visual content.
Outcome:	The proposed model incorporates images that represent Chinese traditional culture, such as famous Chinese traditional paintings, to ensure the authenticity of the Chinese context.

Copied to clipboard

Jialong Wu, Wenbiao Yin, Yong Jiang, Zhenglin Wang, Zekun Xi, Runnan Fang, Linhai Zhang, Yulan He, Deyu Zhou, Pengjun Xie, Fei Huang

Information Retrieval

Challenge:	Large Language Models (LLMs) have demonstrated impressive capabilities across a wide range of natural language processing tasks.
Approach:	They propose a benchmark to assess the ability of LLMs to perform web traversal by using an explore-critic paradigm.
Outcome:	The proposed framework mimics human-like web navigation through an explore-critic paradigm and demonstrates the effectiveness of RAG combined with WebWalker in real-world scenarios.

CPO: Addressing Reward Ambiguity in Role-playing Dialogue via Comparative Policy Optimization (2025.findings-emnlp)

Copied to clipboard

Jing Ye, Rui Wang, Yuchuan Wu, Victor Ma, Feiteng Fang, Fei Huang, Yongbin Li

Dialogue Systems & Conversational Agents

Challenge:	Comparative Policy Optimization (CPO) redefines the reward evaluation paradigm by shifting from sample-wise scoring to comparative group-wise score.
Approach:	They propose a method to optimize subjective tasks by shifting from sample-wise to comparative group-wise scoring.
Outcome:	The proposed framework shifts from sample-wise scoring to comparative group-wise score . it minimizes contextual bias and enables more robust and fair performance evaluation.

ATLAS: Agent Tuning via Learning Critical Steps (2025.findings-acl)

Copied to clipboard

Zhixun Chen, Ming Li, Yuxuan Huang, Yali Du, Meng Fang, Tianyi Zhou

Enhancing Multimodal Unified Representations for Cross Modal Generalization (2025.findings-acl)

Challenge:	Existing agent tuning approaches employ supervised finetuning on entire expert trajectories, but behavior-cloning of full traitories introduces expert bias and weakens generalization to states not covered by the expert data.
Approach:	They propose a method that finetunes LLMs on critical steps in expert trajectories and identifies and finetuns them on these steps with reduced costs.
Outcome:	The proposed method outperforms existing methods and open-source LLM agents on only 30% critical steps in extensive experiments.

Copied to clipboard

Hai Huang, Yan Xia, Shengpeng Ji, Shulei Wang, Hanting Wang, Minghui Fang, Jieming Zhu, Zhenhua Dong, Sashuai Zhou, Zhou Zhao

Extract-Select: A Span Selection Framework for Nested Named Entity Recognition with Generative Adversarial Training (2022.findings-acl)

Challenge:	Existing studies on discrete unified representations overlook important distinctions between different dimensions of features.
Approach:	They propose to use a codebook to optimize unified representations from pretraining and fine- and coarse-grained disentangling to optimize the representations.
Outcome:	The proposed methods improve the interpretability of multimodal unified representations . they use training-free optimization of codebook and fine and coarse cross-modal disentangling .

Copied to clipboard

Peixin Huang, Xiang Zhao, Minghao Hu, Yang Fang, Xinyi Li, Weidong Xiao

Named Entity Recognition

Challenge:	Existing studies treat named entity recognition as a sequential labeling problem.
Approach:	They propose a span selection framework for nested named entity recognition . they propose nesting entities with different input categories would be separately extracted .
Outcome:	The proposed framework outperforms competing models on four benchmark datasets.

Attention Entropy is a Key Factor: An Analysis of Parallel Context Encoding with Full-attention-based Pre-trained Language Models (2025.acl-long)

Copied to clipboard

Zhisong Zhang, Yan Wang, Xinting Huang, Tianqing Fang, Hongming Zhang, Chenlong Deng, Shuaiyi Li, Dong Yu

Universal Information Extraction with Meta-Pretrained Self-Retrieval (2023.findings-acl)

Challenge:	Large language models have demonstrated remarkable performance across a wide range of language tasks due to their remarkable ability in context modeling.
Approach:	They propose to use parallel context encoding to reduce attention entropy by incorporating attention sinks and selective mechanisms to reduce irregular attention . they also propose to incorporate attention sink mechanisms into the parallel encoded context to reduce the irregular attention.
Outcome:	The proposed methods lower irregular attention entropy and narrow performance gaps.

Copied to clipboard

Xin Cong, Bowen Yu, Mengcheng Fang, Tingwen Liu, Haiyang Yu, Zhongkai Hu, Fei Huang, Yongbin Li, Bin Wang

ControlSpeech: Towards Simultaneous and Independent Zero-shot Speaker Cloning and Zero-shot Language Style Control (2025.acl-long)

Challenge:	Existing methods for IE are task-specific, resulting in specialized and isolated approaches for different tasks.
Approach:	They propose a method to retrieve task-specific knowledge from pretrained language models to enhance universal IE by using a Meta-Pretraining Algorithm.
Outcome:	The proposed method achieves the new state-of-the-art on 4 IE tasks, 12 datasets under fully-supervised, low-resource and few-shot scenarios.

Copied to clipboard

Shengpeng Ji, Qian Chen, Wen Wang, Jialong Zuo, Minghui Fang, Ziyue Jiang, Hai Huang, Zehan Wang, Xize Cheng, Siqi Zheng, Zhou Zhao

UniCorn: Towards Self-Improving Unified Multimodal Models through Self-Generated Supervision (2026.acl-long)

Challenge:	Prior zero-shot TTS models only mimic the speaker’s voice without further control and adjustment capabilities while prior controllable TTS systems cannot perform speaker-specific voice generation.
Approach:	They propose a style control module that captures codec representations corresponding to timbre, content, and style in a discrete decoupling codec space.
Outcome:	The proposed system can fully clone the speaker's voice and perform speech-specific adjustment and control functions.

Copied to clipboard

Zhen Fang, Ruiyan Han, XinYu Sun, Yuchen Ma, Ziheng Wang, Yu Zeng, Zehui Chen, Lin Chen, Wenxuan Huang, Wei-Jie Xu, Yi Cao, Feng Zhao

Do Vision-Language Models Have Internal World Models? Towards an Atomic Evaluation (2025.findings-acl)

Challenge:	Unified Multimodal Models have achieved remarkable success in cross-modal comprehension, but a gap persists in their ability to translate internal knowledge into faithful and controllable synthesis.
Approach:	They propose a self-improvement framework that partitions a single UMM into three collaborative roles: Proposer, Solver, and Judge.
Outcome:	The proposed framework improves on TIIF, DPG, CompBench and UniCycle benchmarks.

Copied to clipboard

Qiyue Gao, Xinyu Pi, Kevin Liu, Junrong Chen, Ruolan Yang, Xinqi Huang, Xinyu Fang, Lu Sun, Gautham Kishore, Bo Ai, Stone Tao, Mengyang Liu, Jiaxi Yang, Chao-Jung Lai, Chuanyang Jin, Jiannan Xiang, Benhao Huang, Zeming Chen, David Danks, Hao Su, Tianmin Shu, Ziqiao Ma, Lianhui Qin, Zhiting Hu

KVFKT: A New Horizon in Knowledge Tracing with Attention-Based Embedding and Forgetting Curve Integration (2025.coling-main)

Challenge:	Recent studies have evaluated and shown limitations in specific capabilities such as visual understanding, but a systematic evaluation of VLMs’ fundamental WM abilities remains absent.
Approach:	They propose a framework that assesses perception and prediction to provide an atomic evaluation of VLMs as WMs.
Outcome:	The proposed framework assesses perception and prediction abilities on 15 latest VLMs and compares them to human-level models.

Copied to clipboard

Quanlong Guan, Xiuliang Duan, Kaiquan Bian, Guanliang Chen, Jianbo Huang, Zhiguo Gong, Liangda Fang

Representation Learning

Challenge:	Existing knowledge tracing models do not incorporate forgetting features to improve the learning and answering processes.
Approach:	They propose a new approach in knowledge tracing with attention-based embedding and forgetting curve integration using four real-world datasets to test the model.
Outcome:	The proposed model outperforms the existing knowledge tracing models and eliminates the need for artificial engineering features.

Disentangling Reasoning Logic to Resolve Explicit Knowledge Conflicts (2026.acl-long)

Copied to clipboard

Xianda Zheng, Zijian Huang, Meng-Fen Chiang, Jiamou Liu, Yuan Fang, Michael J. Witbrock, Kaiqi Zhao

Reasoning with OmniThought: A Large CoT Dataset with Verbosity and Cognitive Difficulty Annotations (2026.acl-long)

Challenge:	Existing approaches to resolve explicit knowledge conflicts are based on semantic decoding and auxiliary embedding.
Approach:	They propose a framework that adjudicates conflicts by structuring the underlying logic.
Outcome:	Experiments show that the proposed framework improves on existing models.

Copied to clipboard

Wenrui Cai, Chengyu Wang, Junbing Yan, Jun Huang, Xiangzhong Fang

Reasoning

Challenge:	Existing resources often fail to provide extensive reasoning problems with coherent CoT processes distilled from multiple teacher models.
Approach:	They propose a large-scale dataset featuring 2 million CoT processes generated by multiple powerful LRMs.
Outcome:	The proposed dataset features 2 million CoT processes and is validated by multiple powerful LRMs.

Reverse Preference Optimization for Complex Instruction Following (2025.findings-acl)

Copied to clipboard

Xiang Huang, Ting-En Lin, Feiteng Fang, Yuchuan Wu, Hangyu Li, Yuzhong Qu, Fei Huang, Yongbin Li

MANNER: A Variational Memory-Augmented Model for Cross Domain Few-Shot Named Entity Recognition (2023.acl-long)

Challenge:	Existing methods for identifying and evaluating preference pairs with multiple constraints are noisy.
Approach:	They propose a method that dynamically reverses constraints to ensure the chosen response is perfect.
Outcome:	The proposed method reduces noise in preference pairs by reversing constraints to ensure the chosen response is perfect.

Copied to clipboard

Jinyuan Fang, Xiaobin Wang, Zaiqiao Meng, Pengjun Xie, Fei Huang, Yong Jiang

Low-Resource NLP

Challenge:	Named Entity Recognition (NER) is a fundamental NLP task that aims at classifying mention spans into entity types.
Approach:	They propose a variational memory-augmented few-shot named entity recognition model that uses a memory module to store information from source domain and retrieve relevant information from the memory to augment few-shot task in target domain.
Outcome:	The proposed model can adapt the learned knowledge from source domain to target domain and achieve superior performance on English and Chinese cross domain few-shot NER datasets.

ConvLab-2: An Open-Source Toolkit for Building, Evaluating, and Diagnosing Dialogue Systems (2020.acl-demos)

Copied to clipboard

Qi Zhu, Zheng Zhang, Yan Fang, Xiang Li, Ryuichi Takanobu, Jinchao Li, Baolin Peng, Jianfeng Gao, Xiaoyan Zhu, Minlie Huang

Natural Language Interfaces

Challenge:	ConvLab-2 inherits Convlab's framework but integrates more powerful dialogue models and supports more datasets.
Approach:	They present ConvLab-2, an open-source toolkit that enables researchers to build task-oriented dialogue systems with state-of-the-art models and perform an end-to-end evaluation.
Outcome:	The new tool inherits ConvLab's framework and extends it by integrating many recently proposed state-of-the-art dialogue models.

EAVE: Efficient Product Attribute Value Extraction via Lightweight Sparse-layer Interaction (2024.findings-emnlp)

Copied to clipboard

Li Yang, Qifan Wang, Jianfeng Chi, Jiahao Liu, Jingang Wang, Fuli Feng, Zenglin Xu, Yi Fang, Lifu Huang, Dongfang Liu

Towards Better Entity Linking with Multi-View Enhanced Distillation (2023.acl-long)

Challenge:	Existing methods to extract product attribute value require multiple extractions to obtain all corresponding values.
Approach:	They propose an Efficient product Attribute Value Extraction approach using lightweight sparse-layer interaction.
Outcome:	The proposed method achieves significant efficiency gains with neutral or marginal loss in performance when the context is long and number of attributes is large.

Copied to clipboard

Yi Liu, Yuan Tian, Jianxun Lian, Xinlong Wang, Yanan Cao, Fang Fang, Wen Zhang, Haizhen Huang, Weiwei Deng, Qi Zhang

INT: Establishing Information Transfer for Multilingual Intent Detection and Slot Filling (2025.findings-acl)

Challenge:	Entity linking is a fundamental task in Natural Language Processing (NLP), connecting mentions within unstructured contexts to their corresponding entities in a Knowledge Base (KB).
Approach:	They propose a dual-encoder framework that can efficiently match mentions to two-encoding frameworks by a global-view.
Outcome:	The proposed framework achieves state-of-the-art on several entity linking benchmarks.

Copied to clipboard

Di Wu, Liting Jiang, Bohui Mao, Hongyan Xie, Haoxiang Su, Zhongjiang He, Ruiyu Fang, Shuangyong Song, Hao Huang, Xuelong Li

Sentiment Analysis

Challenge:	Existing studies struggle to achieve performance comparable to that on high-resource languages due to inherent linguistic diversity of multilingual SLU tasks.
Approach:	They propose a multilingual information transfer network to solve these challenges . they propose to reformulate SF as a span prediction problem and introduce a slot-matching attention mechanism to achieve slot alignment across languages.
Outcome:	The proposed model outperforms baseline models on the MASSIVE and MASSIV-UG datasets in overall accuracy across all languages.

MQuinE: a Cure for “Z-paradox” in Knowledge Graph Embedding (2024.emnlp-main)

Copied to clipboard

Yang Liu, Huang Fang, Yunfeng Cai, Mingming Sun

COIG-CQIA: Quality is All You Need for Chinese Instruction Fine-tuning (2025.findings-naacl)

Challenge:	Existing knowledge graph embedding models suffer from Z-paradox, a deficiency in expressiveness . Embedding-based models map each entity and relation into a vector or matrix .
Approach:	They propose a new knowledge graph embedding model that does not suffer from Z-paradox while preserves strong expressiveness to model various relation patterns with theoretical justification.
Outcome:	The proposed model outperforms existing models on link prediction tasks while maintaining strong expressiveness.

Copied to clipboard

Yuelin Bai, Xeron Du, Yiming Liang, Leo Jin, Junting Zhou, Ziqiang Liu, Feiteng Fang, Mingshan Chang, Tianyu Zheng, Xincheng Zhang, Nuo Ma, Zekun Moore Wang, Ruibin Yuan, Haihong Wu, Hongquan Lin, Wenhao Huang, Jiajun Zhang, Chenghua Lin, Jie Fu, Min Yang, Shiwen Ni, Ge Zhang

MeetDot: Videoconferencing with Live Translation Captions (2021.emnlp-demo)

Challenge:	Existing datasets for Chinese instruction tuning are not well-aligned with Chinese users’ interaction patterns.
Approach:	They propose to use Chinese instruction tuning datasets to improve instruction fine-tuning for Chinese users.
Outcome:	The proposed dataset shows that Chinese models achieve competitive performance in diverse benchmarks.

Copied to clipboard

Arkady Arkhangorodsky, Christopher Chu, Scot Fang, Yiqi Huang, Denglin Jiang, Ajay Nagesh, Boliang Zhang, Kevin Knight

Text Generation

Challenge:	MeetDot is a videoconferencing system with live translation captions overlaid on screen . currently, the system supports speech and captions in 4 languages .
Approach:	They propose a videoconferencing system with live translation captions overlaid on screen . the system supports speech and captions in 4 languages and combines automatic speech recognition and machine translation in a cascade .
Outcome:	The proposed system supports speech and captions in 4 languages and has very tight latency requirements to have acceptable call quality.

i-Code Studio: A Configurable and Composable Framework for Integrative AI (2024.emnlp-demo)

Copied to clipboard

Yuwei Fang, Mahmoud Khademi, Chenguang Zhu, Ziyi Yang, Reid Pryzant, Yichong Xu, Yao Qian, Takuya Yoshioka, Lu Yuan, Michael Zeng, Xuedong Huang

CATS: A Pragmatic Chinese Answer-to-Sequence Dataset with Large Scale and High Quality (2023.acl-long)

Challenge:	Existing frameworks for Integrative AI lack flexibility and composability to handle multimodal tasks.
Approach:	They propose a configurable framework for Integrative AI that orchestrates multiple pre-trained models to conduct complex multimodal tasks.
Outcome:	The proposed framework achieves impressive results on zero-shot multimodal tasks . it can communicate and personalize for users, and it can be used in a multimodal agent .

Copied to clipboard

Liang Li, Ruiying Geng, Chengyang Fang, Bing Li, Can Ma, Rongyu Cao, Binhua Li, Fei Huang, Yongbin Li

Structured Data in NLP

Challenge:	Current datasets bias in the English language while leaving other languages underexplored.
Approach:	They propose a Chinese answer-to-sequence dataset with high quality and large scale . they propose encoding space for two hybrid knowledge resources to convert this task to a graph-totext problem.
Outcome:	The proposed method is effective in generating textual descriptions for the Chinese answer-to-sequence dataset.

SynWorld: Virtual Scenario Synthesis for Agentic Action Knowledge Refinement (2025.acl-short)

Copied to clipboard

Runnan Fang, Xiaobin Wang, Yuan Liang, Shuofei Qiao, Jialong Wu, Zekun Xi, Ningyu Zhang, Yong Jiang, Pengjun Xie, Fei Huang, Huajun Chen

Challenge:	Using Large Language Models (LLMs)-based agents can enhance their understanding of environments and tasks.
Approach:	They propose a framework that allows agents to synthesize possible scenarios with multi-step action invocation within the action space and perform Monte Carlo Tree Search exploration to refine their action knowledge in the current environment.
Outcome:	The proposed framework synthesizes possible scenarios with multi-step action invocation within the action space and performs Monte Carlo Tree Search exploration to refine action knowledge in the current environment.

LIME: Less Is More for MLLM Evaluation (2025.findings-acl)

Copied to clipboard

King Zhu, Qianbo Zang, Shian Jia, Siwei Wu, Feiteng Fang, Yizhi Li, Shuyue Guo, Tianyu Zheng, Jiawei Guo, Bo Li, Haoning Wu, Xingwei Qu, Jian Yang, Ruibo Liu, Xiang Yue, Jiaheng Liu, Chenghua Lin, Hamid Alinejad-Rokny, Min Yang, Shiwen Ni, Wenhao Huang, Ge Zhang

ProMed: Shapley Information Gain Guided Reinforcement Learning for Proactive Medical LLMs (2026.acl-long)

Challenge:	Existing MLLM benchmarks and unified evaluation frameworks cannot accurately and efficiently reflect the ability of MLMLs.
Approach:	They propose a semi-automated benchmark curated using a pipeline that filters out uninformative samples and eliminates answer leakage by focusing on tasks that require image-based understanding.
Outcome:	The proposed benchmark reduces the number of samples by 76% and evaluation time by 77% while it can more effectively distinguish different models’ abilities.

Copied to clipboard

Hongxin Ding, Baixiang Huang, Yue Fang, Weibin Liao, Xinke Jiang, Jinyang Zhang, Yinghao Zhu, Zheng Li, Liantao Ma, Junfeng Zhao, Yasha Wang

Challenge:	Existing medical Large Language Models (LLMs) follow a reactive paradigm, risking diagnostic errors by answering before seeking sufficient details.
Approach:	They propose a reinforcement learning framework that transitions LLMs toward a proactive paradigm, enabling them to ask clinically valuable questions before decision-making.
Outcome:	Experiments on partial-information medical benchmarks show that ProMed outperforms state-of-the-art methods by 6.29% on average and delivers a 54.45% gain over the reactive paradigm.

Deep Research with Open-Domain Evaluation and Multi-Stage Guardrails for Safety (2026.acl-long)

Copied to clipboard

Wei-Chieh Huang, Henry Peng Zou, Yaozu Wu, Dongyuan Li, Yankai Chen, Weizhi Zhang, Yangning Li, Angelo Zangari, Jizhou Guo, Chunyu Miao, Liancheng Fang, Langzhou He, Yinghui Li, Renhe Jiang, Philip S. Yu

Multi-Agent Autonomous Driving Systems with Large Language Models: A Survey of Recent Advances, Resources, and Future Directions (2025.findings-emnlp)

Challenge:	Existing deep research frameworks lack adequate evaluation procedures and stage-specific protections.
Approach:	They propose a framework with open-domain evaluation and a stage-wise safety benchmark to address this oversight.
Outcome:	The proposed framework improves defense success rates by 16.53% while reducing over-refusal rates to approximately 6%.

Copied to clipboard

Yaozu Wu, Dongyuan Li, Yankai Chen, Renhe Jiang, Henry Peng Zou, Wei-Chieh Huang, Yangning Li, Liancheng Fang, Zhen Wang, Philip S. Yu

Knowledge Injection Exists in MoE? Exploring Expert-Aware Contrast Decoding in MoE for Mitigating LLMs’ Hallucinations (2026.acl-long)

Challenge:	Large Language Models (LLMs) are used to assist with driving decisions, but they face limitations in perception and computational demands.
Approach:	They propose a survey of LLM-based multi-agent ADSs and their applications . they analyze agent-human interactions in scenarios where LLM agents engage with humans .
Outcome:	The proposed approach reduces human intervention and improves safety and efficiency.

Copied to clipboard

Xinyue Fang, Zhiliang Tian, Zhen Huang, Ziyi Pan, Zhihua Wen, Xi Wang, Quntian Fang, Dongsheng Li

OmniAlign-V: Towards Enhanced Alignment of MLLMs with Human Preference (2025.acl-long)

Challenge:	Existing methods to mitigate hallucinations include prompt engineering and model optimization, but lack domain generalization and potential errors in fine-tuning data may exacerbate the hallucism.
Approach:	They propose an expert-aware adaptive contrast decoding that uses expert differences in MoE’s higher layers to mitigate hallucinations on QA tasks.
Outcome:	The proposed method outperforms baseline models on four datasets Large language models (LLMs) show strong performance but suffer from hallucinations, limiting their application.

Copied to clipboard

Xiangyu Zhao, Shengyuan Ding, Zicheng Zhang, Haian Huang, Maosongcao Maosongcao, Jiaqi Wang, Weiyun Wang, Xinyu Fang, Wenhai Wang, Guangtao Zhai, Hua Yang, Haodong Duan, Kai Chen