Papers by Wenhao Li

WeCheck: Strong Factual Consistency Checker via Weakly Supervised Learning (2023.acl-long)

Challenge:	Existing methods for named entity recognition classify mentions into fixed set of predefined entity types but in many real-world scenarios, new entity types are incrementally involved.
Approach:	They propose a two-stage framework Learn-and-Review for continual named entity recognition to alleviate inter-type confusion.
Outcome:	The proposed framework outperforms the state-of-the-art method on CoNLL-03 and OntoNotes-5.0.

Copied to clipboard

Wenhao Wu, Wei Li, Xinyan Xiao, Jiachen Liu, Sujian Li, Yajuan Lyu

STARS: A Unified Framework for Singing Transcription, Alignment, and Refined Style Annotation (2025.findings-acl)

Challenge:	Existing factual consistency metrics are often uncontrollably generating text that is factually inconsistent with inputs.
Approach:	They propose a weakly supervised framework that is directly trained on actual generated samples from language models with weakly annotated labels.
Outcome:	The proposed framework improves on the TRUE benchmark by 3.3% over existing methods with 435M parameters.

Copied to clipboard

Wenxiang Guo, Yu Zhang, Changhao Pan, Zhiyuan Zhu, Ruiqi Li, ZheTao Chen, Wenhao Xu, Fei Wu, Zhou Zhao

Speech & Audio in NLP

Challenge:	Existing automated singing annotation (ASA) methods tackle isolated aspects of the annotation pipeline.
Approach:	They propose a framework that addresses transcription, alignment, and refined style annotations.
Outcome:	The proposed framework delivers comprehensive multi-level annotations encompassing: (1) precise phoneme-audio alignment, (2) robust note transcription and temporal localization, (3) expressive vocal technique identification, and (4) global stylistic characterization including emotion and pace.

LIFTED: Multimodal Clinical Trial Outcome Prediction via Large Language Models and Mixture-of-Experts (2025.findings-emnlp)

Copied to clipboard

Wenhao Zheng, Liaoyaqi Wang, Dongshen Peng, Hongxia Xu, Yun Li, Hongtu Zhu, Tianfan Fu, Huaxiu Yao

Curriculum-RLAIF: Curriculum Alignment with Reinforcement Learning from AI Feedback (2026.findings-acl)

Challenge:	Clinical trials are costly and pivotal processes that require substantial expenses . a new approach to integrate multimodal data for clinical outcome prediction is needed .
Approach:	a proposed framework transforms modality-specific data into natural language descriptions . a sparse Mixture-of-Experts mechanism then identifies shared patterns across modalities .
Outcome:	a proposed framework outperforms baseline methods in predicting clinical trial outcomes . it transforms modality-specific data into natural language descriptions, encoded via unified encoders .

Copied to clipboard

Jiaye Lin, Mengdi Li, Xufeng Zhao, Wenhao Lu, Peilin Zhao, Stefan Wermter, Di Wang

Challenge:	Existing approaches to align large language models with human preferences are limited in generalizability due to distribution shift, preference label noise, and mismatch of challenging samples with model capacity.
Approach:	They propose a framework that constructs preference pairs with varying difficulty levels and then produces a specific curriculum for reward model training.
Outcome:	The proposed framework improves generalizability of reward models by a significant margin without incurring additional inference costs compared to existing non-curriculum baselines.

AgentBank: Towards Generalized LLM Agents via Fine-Tuning on 50000+ Interaction Trajectories (2024.findings-emnlp)

Copied to clipboard

Yifan Song, Weimin Xiong, Xiutian Zhao, Dawei Zhu, Wenhao Wu, Ke Wang, Cheng Li, Wei Peng, Sujian Li

Leveraging Dual Process Theory in Language Agent Framework for Real-time Simultaneous Human-AI Collaboration (2025.acl-long)

Challenge:	Existing studies focus on specialized agents designed for particular tasks.
Approach:	They propose to scale annotated interaction trajectories and fine-tune LLMs on AgentBank to get a series of agent models, Samoyed.
Outcome:	The proposed model can scale to get generalized agent capabilities.

Copied to clipboard

Shao Zhang, Xihuai Wang, Wenhao Zhang, Chaoran Li, Junru Song, Tingyu Li, Lin Qiu, Xuezhi Cao, Xunliang Cai, Wen Yao, Weinan Zhang, Xinbing Wang, Ying Wen

Linguistic Theories

Challenge:	Large language models (LLMs) excel in turn-by-turn human-AI collaboration but struggle with simultaneous tasks requiring real-time interaction.
Approach:	They propose a language agent framework that integrates System 1 and System 2 for efficient real-time simultaneous human-AI collaboration.
Outcome:	The proposed framework improves on existing LLM-based agents and human collaborators by integrating Theory of Mind and asynchronous reflection to infer human intentions and perform reasoning-based autonomous decisions.

Aligning Large Language Models with Human Preferences through Representation Engineering (2024.acl-long)

Copied to clipboard

Wenhao Liu, Xiaohua Wang, Muling Wu, Tianlong Li, Changze Lv, Zixuan Ling, Zhu JianHao, Cenyuan Zhang, Xiaoqing Zheng, Xuanjing Huang

Exploring In-Context Learning for Knowledge Grounded Dialog Generation (2023.findings-emnlp)

Challenge:	Existing methods for achieving this alignment involve employing reinforcement learning from human feedback (RLHF) Existing approaches involve using RLHF to fine-tune LLMs based on human labels . however, RLRF is susceptible to instability during fine- tuning and presents challenges in implementation.
Approach:	They propose to use reinforcement learning from human feedback to fine-tune large language models with human preferences to achieve precise control of model behavior.
Outcome:	Experiments show that RAHF can be used to capture and manipulate representations to align with a broad spectrum of human preferences or values rather than being confined to a single concept or function.

Copied to clipboard

Qinyu Chen, Wenhao Wu, Sujian Li

Dialogue Systems & Conversational Agents

Challenge:	Existing knowledge grounded dialog generation models are prone to hallucination and produce factually inaccurate outputs.
Approach:	They propose a retrieval-based framework which leverages in-context learning and retrieval techniques to enhance LLMs on knowledge grounded dialog generation.
Outcome:	The proposed framework outperforms existing training-based models on a large-scale knowledge graph with 1M+ facts and is expected to perform knowledge-intensive tasks.

Automatic Poetry Generation with Mutual Reinforcement Learning (D18-1)

Copied to clipboard

Xiaoyuan Yi, Maosong Sun, Ruoyu Li, Wenhao Li

Text Generation

Challenge:	Existing models for automatic poetry generation are based on maximum likelihood estimation (MLE) MLE-based models tend to remember common patterns of the poetry corpus, which results in loss-evaluation mismatch.
Approach:	They propose to model the criteria and use them as explicit rewards to guide gradient update by reinforcement learning to motivate the model to pursue higher scores.
Outcome:	The proposed model outperforms the current state-of-the-art model and improves on Chinese poetry.

AutoScraper: A Progressive Understanding Web Agent for Web Scraper Generation (2024.emnlp-main)

Copied to clipboard

Wenhao Huang, Zhouhong Gu, Chenghao Peng, Jiaqing Liang, Zhixu Li, Yanghua Xiao, Liqian Wen, Zulong Chen

Challenge:	Existing methods for web scraping suffer from limited adaptability and scalability when faced with a new website.
Approach:	They propose a framework that generates web scrapers with large language models and a new executability metric to measure the performance of web scraper generation tasks.
Outcome:	The proposed framework can handle diverse web environments more efficiently.

Parameter Efficient Multi-task Fine-tuning by Learning to Transfer Token-wise Prompts (2023.findings-emnlp)

Copied to clipboard

Muling Wu, Wenhao Liu, Jianhan Xu, Changze Lv, Zixuan Ling, Tianlong Li, Longtao Huang, Xiaoqing Zheng, Xuanjing Huang

Document Segmentation Matters for Retrieval-Augmented Generation (2025.findings-acl)

Challenge:	Prompt tuning has been proven to be successful on various tasks by incorporating a small number of trainable parameters while freezing large pre-trained language models.
Approach:	They propose a token-wise prompt tuning method that uses a bank of finer-grained soft prompt tokens to generate an instance-dependent prompt.
Outcome:	The proposed method performs far better than full parameter fine-tuned models and achieves state-of-the-art by tuning only 0.035% parameters on 14 datasets.

Copied to clipboard

Zhitong Wang, Cheng Gao, Chaojun Xiao, Yufei Huang, Shuzheng Si, Kangyang Luo, Yuzhuo Bai, Wenhao Li, Tangjian Duan, Chuancheng Lv, Guoshan Lu, Gang Chen, Fanchao Qi, Maosong Sun

Syntactic Text Processing

Challenge:	Existing rule-based chunking methods lead to suboptimal splits, where overly large chunks introduce irrelevant information and small chunks lack semantic coherence.
Approach:	They propose a method that leverages document summaries as pseudo-instructions to guide chunking by computing semantic similarity between sentences and the summary.
Outcome:	Experiments on multiple open-domain question-answering benchmarks show that PIC significantly improves retrieval accuracy (Hits@k) and end-to-end QA performance (Exact Match) without any additional training.

Polynomial Expansion Rank Adaptation: Enhancing Low-Rank Fine-Tuning with High-Order Interactions (2026.findings-acl)

Copied to clipboard

Wenhao Zhang, Lin Mu, Li Ni, Peiquan Jin, Yiwen Zhang

Evade the Trap of Mediocrity: Promoting Diversity and Novelty in Text Generation via Concentrating Attention (2022.emnlp-main)

Challenge:	Low-rank adaptation (LoRA) is a widely used strategy for efficient fine-tuning of large language models, but its strictly linear structure limits expressive capacity.
Approach:	They propose a method that introduces structured polynomial expansion directly into the low-rank factor space.
Outcome:	The proposed method outperforms state-of-the-art methods across diverse benchmarks.

Copied to clipboard

Wenhao Li, Xiaoyuan Yi, Jinyi Hu, Maosong Sun, Xing Xie

STORM-BORN: A Challenging Mathematical Derivations Dataset Curated via a Human-in-the-Loop Multi-Agent Framework (2025.findings-acl)

Challenge:	Recent studies have shown that powerful Transformer architectures produce dull high-frequency phrases, severely hurting the diversity and novelty of generated text.
Approach:	They propose a method to control the sharpness of the attention distribution by python code and use it to learn a Bayesian approximation of posterior attention.
Outcome:	The proposed method improves diversity and novelty while maintaining comparable quality on conditional and unconditional generation tasks.

Copied to clipboard

Wenhao Liu, Zhenyi Lu, Xinyu Hu, Jerry Zhang, Dailin Li, Jiacheng Cen, Huilin Cao, Haiteng Wang, Yuhan Li, Xie Kun, Dandan Li, Pei Zhang, Chengbo Zhang, Yuxiang Ren, Xiaohong Huang, Yan Ma

Reasoning

Challenge:	Existing datasets suffer from outdated and insufficient challenging content, neglecting human-like reasoning, and limited reliability due to single-LLM generation.
Approach:	They propose a human-in-the-loop, multi-agent data generation framework that integrates reasoning-dense filters, multiagent collaboration, and human mathematicians’ evaluations to ensure the reliability and quality of the dataset.
Outcome:	The proposed framework improves accuracy and quality of the 2,000-synthesized datasets by integrating reasoning-dense filters, multi-agent collaboration, and human mathematicians’ evaluations.

More Tokens, Lower Precision: Towards the Optimal Token-Precision Trade-off in KV Cache Compression (2025.findings-emnlp)

Copied to clipboard

Jiebin Zhang, Dawei Zhu, Yifan Song, Wenhao Wu, Chuqiao Kuang, Xiaoguang Li, Lifeng Shang, Qun Liu, Sujian Li

Challenge:	storing more tokens in the KV cache at lower precision can enhance the long-context performance of large language models.
Approach:	They propose a token-precision trade-off strategy to optimize KV cache compression . they also propose storing more tokens in the KV at lower precision .
Outcome:	The proposed method achieves an optimal point within the Information Bottleneck compared to standalone KV pruning or KV quantization.

On-policy Reinforcement Fine-tuning with Offline reward for Multi-step Embodied Planning (2026.acl-long)

Copied to clipboard

Di Wu, Jiaxin Fan, Chloe Gu, Guanbo Wang, Wei Yin, Wenhao Li, Bo Jin

Tell Me What You Don’t Know: Enhancing Refusal Capabilities of Role-Playing Agents via Representation Space Analysis and Editing (2025.findings-acl)

Challenge:	Embodied planning requires agents to make coherent multi-step decisions based on dynamic visual observations and verbal goals.
Approach:	They propose an On-policy Reinforcement fine-tuning framework with offline rewards for Embodied Task Planning that preserves generalization benefits of RFT while addressing costly interaction and sparse rewards.
Outcome:	The proposed framework outperforms closed-source and online-RL methods on EmbodiedBench, a recent benchmark for interactive embodied tasks.

Copied to clipboard

Wenhao Liu, Siyu An, Junru Lu, Muling Wu, Tianlong Li, Xiaohua Wang, Changze Lv, Xiaoqing Zheng, Di Yin, Xing Sun, Xuanjing Huang

Representation Learning

Challenge:	Role-playing Agents (RPAs) struggle to recognize and respond to hard queries that conflict with their role-play knowledge.
Approach:	They propose a lightweight representation editing approach that conveniently shifts conflicting requests to the rejection region, thereby enhancing the model’s refusal accuracy.
Outcome:	The proposed model improves RPAs’ refusal ability of conflicting requests while maintaining their general role-playing capabilities.

RECALL: REpresentation-aligned Catastrophic-forgetting ALLeviation via Hierarchical Model Merging (2025.emnlp-main)

Copied to clipboard

Bowen Wang, Haiyuan Wan, Liwen Shi, Chen Yang, Peng He, Yue Ma, Haochen Han, Wenhao Li, Tiao Tan, Yongjian Li, Fangming Liu, Gong Yifan, Sheng Zhang

Agentic Episodic Control (2026.findings-acl)

Challenge:	Existing models that require task labels or performance trade-offs are susceptible to catastrophic forgetting.
Approach:	They propose a representation-aware model merging framework for continual learning without access to historical data.
Outcome:	The proposed framework outperforms baselines in knowledge retention and generalization across five NLP tasks and multiple continual learning scenarios.

Copied to clipboard

Xidong Yang, Wenhao Li, Junjie Sheng, Yun Hua, Haosheng Chen, Chuyun Shen, Xiangfeng Wang

Teaching According to Talents! Instruction Tuning LLMs with Competence-Aware Curriculum Learning (2025.findings-emnlp)

Challenge:	Existing methods for reinforcement learning (RL) are limited by poor data efficiency and weak generalization.
Approach:	They propose a novel architecture that integrates large language models into episodic RL.
Outcome:	The proposed architecture achieves 2–6 higher data efficiency than baselines and is the only method to solve complex tasks like UnlockLocal with over 90% success.

Copied to clipboard

Yangning Li, Tingwei Lu, Yinghui Li, Yankai Chen, Wei-Chieh Huang, Wenhao Jiang, Hui Wang, Hai-Tao Zheng, Philip S. Yu

VideoCuRL: Video Curriculum Reinforcement Learning with Orthogonal Difficulty Decomposition (2026.acl-long)

Challenge:	Efficient instruction tuning aims to enhance the ultimate performance of large language models (LLMs) current methods suffer from the curriculum rigidity, resulting in a fixed and potentially sub-optimal learning trajectory.
Approach:	a framework for efficient instruction tuning is proposed to address the issue of curriculum rigidity . current methods rely on static heuristic difficulty metrics and fail to adapt to evolving capabilities .
Outcome:	Efficient instruction tuning aims to enhance the ultimate performance of large language models . current methods suffer from the curriculum rigidity, resulting in a fixed learning trajectory .

Copied to clipboard

Hongbo Jin, Kuanwei Lin, Wenhao Zhang, Yichen Jin, Ge Li

MKT: A Multi-Stage Knowledge Transfer Framework to Mitigate Catastrophic Forgetting in Multi-Domain Chinese Spelling Correction (2025.emnlp-industry)

Challenge:	Reinforcement Learning (RL) is crucial for Video-LLMs with complex spatiotemporal reasoning.
Approach:	They propose a framework that decomposes difficulty into two axes in video understanding . they employ efficient, training-free proxies to map data onto a 2D curriculum grid .
Outcome:	The proposed framework surpasses strong RL baselines on reasoning and perception tasks.

Copied to clipboard

Peng Xing, Yinghui Li, Shirong Ma, Xinnian Liang, Haojing Huang, Yangning Li, Shu-Yu Guo, Hai-Tao Zheng, Wenhao Jiang, Ying Shen

AutoPKG: An Automated Framework for Dynamic E-commerce Product-Attribute Knowledge Graph Construction (2026.findings-acl)

Challenge:	Chinese Spelling Correction (CSC) is a model that detects and corrects spelling errors in given sentences.
Approach:	They propose a model-agnostic model with an evolving teacher model and dynamic distillation weights for knowledge transfer in each domain rather than focusing solely on new domain knowledge.
Outcome:	The proposed model-agnostic framework is based on an evolving teacher model and dynamic distillation weights for knowledge transfer in each domain, rather than focusing solely on new domain knowledge.

Copied to clipboard

Pollawat Hongwimol, Haoning Shang, Chutong Wang, Zhichao Wan, Yi Gao, Yuanming Li, Lin Gui, Wenhao Sun, Cheng Yu

LLaMAX: Scaling Linguistic Horizons of LLM by Enhancing Translation Capabilities Beyond 100 Languages (2024.findings-emnlp)

Challenge:	Product attribute extraction in e-commerce is bottlenecked by ontologies that are inconsistent, incomplete, and costly to maintain.
Approach:	They propose a multi-agent Large Language Model framework that constructs a Product-attribute Knowledge Graph from multimodal product content.
Outcome:	The proposed framework achieves 0.953 WKE for product types, 0.724 WKEs for attribute keys, and 0.531 edge-level accuracy for value assertions after canonicalization on a large real-world marketplace catalog dataset from Lazada (Alibaba).

Copied to clipboard

Yinquan Lu, Wenhao Zhu, Lei Li, Yu Qiao, Fei Yuan

Multilinguality

Challenge:	Large Language Models (LLMs) exhibit exceptional translation capabilities in high-resource language tasks, yet their effectiveness in low-resourced languages is suboptimal.
Approach:	They conduct extensive multilingual continual pre-training on the LLaMA series models and develop LLiMAX for translation support across more than 100 languages.
Outcome:	The proposed model achieves higher translation performance than existing open-source models and performs on-par with specialized translation model on the Flores-101 benchmark.

FRSUM: Towards Faithful Abstractive Summarization via Enhancing Factual Robustness (2022.findings-emnlp)

Copied to clipboard

Wenhao Wu, Wei Li, Jiachen Liu, Xinyan Xiao, Ziqiang Cao, Sujian Li, Hua Wu

Robustness in NLP

Challenge:	Existing models of abstractive summarization are able to generate fluent and coherent summaries, but they still suffer from the unfaithful generation problem.
Approach:	They propose to improve the faithfulness of existing models by enhancing their factual robustness by using a novel training strategy, namely FRSUM, which teaches the model to defend against both explicit adversarial samples and implicit factual adversarials.
Outcome:	The proposed training strategy improves faithfulness of various models, such as T5, BART, and T5 .

Rethinking the Roles of Large Language Models in Chinese Grammatical Error Correction (2025.acl-industry)

Copied to clipboard

Yinghui Li, Shang Qin, Jingheng Ye, Haojing Huang, Yangning Li, Shu-Yu Guo, Libo Qin, Xuming Hu, Wenhao Jiang, Hai-Tao Zheng, Philip S. Yu

MEIC-DT: Memory-Efficient Incremental Clustering for Long-Text Coreference Resolution with Dual-Threshold Constraints (2026.findings-acl)

Challenge:	Recent studies have shown that Large Language Models’ performance as correctors on Chinese Grammatical Error Correction (CGEC) remains unsatisfactory due to the challenging nature of the task.
Approach:	They propose a training framework EXAM that uses LLMs as explainers to enhance CGEC small models and a novel evaluation method SEE that utilizes LLM as evaluators to bring more reasonable evaluations.
Outcome:	The proposed methods improve the performance of LLMs on Chinese Grammatical Error Correction (CGEC) task.

Copied to clipboard

Kangyang Luo, Shuzheng Si, Yuzhuo Bai, Cheng Gao, Zhitong Wang, Cheng Huang, Yingli Shen, Yufeng Han, Wenhao Li, Cunliang Kong, Maosong Sun

CIF-Bench: A Chinese Instruction-Following Benchmark for Evaluating the Generalizability of Large Language Models (2024.findings-acl)

Challenge:	Existing supervised neural methods are underexplored for coreference resolution, especially in incremental clustering.
Approach:	They propose a dual-threshold incremental clustering approach based on a lightweight Transformer.
Outcome:	Experiments on common benchmarks show that MEIC-DT achieves highly competitive coreference performance under stringent memory constraints.

Copied to clipboard

Yizhi Li, Ge Zhang, Xingwei Qu, Jiali Li, Zhaoqun Li, Noah Wang, Hao Li, Ruibin Yuan, Yinghao Ma, Kai Zhang, Wangchunshu Zhou, Yiming Liang, Lei Zhang, Lei Ma, Jiajun Zhang, Zuowen Li, Wenhao Huang, Chenghua Lin, Jie Fu

Improving Cross-lingual Transfer with Contrastive Negative Learning and Self-training (2024.lrec-main)

Challenge:	a recent study shows that large language models have limited generalization in low-resource languages like Chinese.
Approach:	They propose to evaluate the zero-shot generalizability of large language models to the Chinese language . they release only half of the dataset publicly, with the remainder kept private .
Outcome:	The Chinese Instruction-Following Benchmark evaluates the generalizability of LLMs to the Chinese language.

Copied to clipboard

Guanlin Li, Xuechen Zhao, Amir Jafari, Wenhao Shao, Reza Farahbakhsh, Noel Crespi

Cross-Lingual Transfer

Challenge:	Recent studies improve cross-lingual transfer learning by better aligning the internal representations within the multilingual model or exploring the information of the target language using self-training.
Approach:	They propose to use negative pairs to align the multilingual model and self-train the model to converge on the obtained clean pseudo-labels.
Outcome:	The proposed method improves upon the baseline models and can serve as a beneficial complement to the alignment-based methods.

Is There a One-Model-Fits-All Approach to Information Extraction? Revisiting Task Definition Biases (2024.findings-emnlp)

Copied to clipboard

Wenhao Huang, Qianyu He, Zhixu Li, Jiaqing Liang, Yanghua Xiao

Lego-MT: Learning Detachable Models for Massively Multilingual Machine Translation (2023.findings-acl)

Challenge:	Definition bias is a negative phenomenon that can mislead models.
Approach:	They propose a framework that measures definition bias, bias-aware fine-tuning and task-specific bias mitigation to mitigate definition bias in information extraction.
Outcome:	The proposed framework mitigates definition bias in information extraction tasks by measuring definition bias, bias-aware fine-tuning, and task-specific bias mitigation.

Copied to clipboard

Fei Yuan, Yinquan Lu, Wenhao Zhu, Lingpeng Kong, Lei Li, Yu Qiao, Jingjing Xu

Multilinguality

Challenge:	Existing monolithic models for multilingual neural machine translation encounter parameter interference and inefficient inference for large models.
Approach:	They propose a detachable multi-way model that assigns each language to an individual branch . they use data from OPUS to build a translation benchmark covering 433 languages .
Outcome:	The proposed model outperforms existing models in OPUS and is faster than existing models.

Precisely the Point: Adversarial Augmentations for Faithful and Informative Text Generation (2022.emnlp-main)

Copied to clipboard

Wenhao Wu, Wei Li, Jiachen Liu, Xinyan Xiao, Sujian Li, Yajuan Lyu

Robustness in NLP

Challenge:	Existing models for text generation are weak enough to handle perturbations in inputs, leading to degeneration in faithfulness and informativeness.
Approach:	They propose a framework for improving faithfulness and informativeness of Seq2Seq models by perturbing word representations and word swapping.
Outcome:	The proposed framework improves faithfulness and informativeness of Seq2Seq models under automatic and human evaluation settings.

Efficient Transformer Parameter Reuse via Zero-Token Mechanism (2026.findings-acl)

Copied to clipboard

Guanghao Li, Wenhao Jiang, Li Shen, Ming Tang, Chun Yuan

Enhanced Reasoning for Biomedical Document-Level Relation Extraction via a Novel Cascade Language Model Framework (2026.acl-long)

Challenge:	Existing approaches to scaling up parameter counts are impractical for users with limited computational resources.
Approach:	They propose a decoupled parameter cycling strategy that employs a head-tail decoupling strategy to decouple the first (head) and last (tail) layers from the parameter cycling process.
Outcome:	The proposed approach achieves superior performance under strict parameter constraints and significantly reduces computational overhead via early exits.

Copied to clipboard

Haohua Song, Wenhao Gu, Zhijing Li, null Yunwenyu, Tiantian Zhu, Xiao Yang, Zexuan Zhu

CityCube: Benchmarking Cross-view Spatial Reasoning on Vision-Language Models in Urban Environments (2026.acl-long)

Challenge:	Pre-trained language models (PLMs) are the leading paradigm in document-level relation extraction.
Approach:	They propose a cascade framework that leverages the complementary strengths of PLMs and LLMs through a detect-then-rethink paradigm.
Outcome:	The proposed framework improves on BioRED and CDR datasets and improves existing models.

Copied to clipboard

Haotian Xu, Yue Hu, Zhengqiu Zhu, Chen Gao, Ziyou Wang, Junreng Rao, Wenhao Lu, Weishi Li, Quanjun Yin, Yong Li

CriticLean: Critic-Guided Reinforcement Learning for Mathematical Formalization (2026.acl-long)

Challenge:	Existing benchmarks focus on indoor or street settings, overlooking challenges of open-ended urban spaces.
Approach:	They propose a benchmark to probe cross-view spatial reasoning capabilities of current VLMs in urban settings.
Outcome:	The citycube benchmark examines the performance of current vision-language models in urban environments.

Copied to clipboard

Zhongyuan Peng, Yifan Yao, Kaijing Ma, Shuyue Guo, Yizhe Li, Yichi Zhang, Chenchen Zhang, Yifan Zhang, Zhouliang Yu, Luming Li, Minghao Liu, Yihang Xia, Jiawei Shen, Yuchen Wu, Yixin Cao, Zhaoxiang Zhang, Wenhao Huang, Jiaheng Liu, Ge Zhang

Reasoning

Challenge:	Existing approaches to formalizing mathematical statements face limitations in accuracy, especially in the context of complex, highlevel problems that involve sophisticated mathematical reasoning.
Approach:	They propose a CriticLean framework that elevates the role of the critic from a passive validator to an active learning component and introduce a benchmark to measure models’ ability to distinguish semantically correct from incorrect formalizations.
Outcome:	The proposed framework outperforms open- and closed-source benchmarks and shows that it significantly outperformed existing models.

ChatMusician: Understanding and Generating Music Intrinsically with LLM (2024.findings-acl)

Copied to clipboard

Ruibin Yuan, Hanfeng Lin, Yi Wang, Zeyue Tian, Shangda Wu, Tianhao Shen, Ge Zhang, Yuhang Wu, Cong Liu, Ziya Zhou, Liumeng Xue, Ziyang Ma, Qin Liu, Tianyu Zheng, Yizhi Li, Yinghao Ma, Yiming Liang, Xiaowei Chi, Ruibo Liu, Zili Wang, Chenghua Lin, Qifeng Liu, Tao Jiang, Wenhao Huang, Wenhu Chen, Jie Fu, Emmanouil Benetos, Gus Xia, Roger Dannenberg, Wei Xue, Shiyin Kang, Yike Guo

InstructEval: Instruction-Tuned Text Evaluator from Human Preference (2024.findings-acl)

Challenge:	Despite LLMs' impressive capabilities in musical knowledge, music reasoning remains an unsolved task.
Approach:	They propose an open-source large language model (LLM) that integrates intrinsic musical abilities into LLaMA2 and GPT-3.5.
Outcome:	The proposed model can understand and generate music with a pure text tokenizer without external multi-modal neural structures or tokenizers.

Copied to clipboard

Wenhao Wu, Wei Li, Xinyan Xiao, Jiachen Liu, Sujian Li

ReStyle-TTS: Relative and Continuous Style Control for Zero-Shot Speech Synthesis (2026.findings-acl)

Challenge:	InstructEval is a general text evaluator based on open-source Large Language Models (LLMs).
Approach:	They propose to build a general multi-aspect text evaluator based on open-source Large Language Models (LLMs) they use extensive open Human Preference Modeling datasets and a small set of multi-spect annotated data to overcome the shortage of annotation resources for multi-task evaluations.
Outcome:	The proposed model performs comparable or superior to commercial LLMs like ChatGPT or GPT-4 in terms of both overall and multi-aspect evaluation tasks.

Copied to clipboard

Haitao Li, Chunxiang Jin, Chenglin Li, Wenhao Guan, Zhengxing Huang, Xie Chen

GLTW: Joint Improved Graph Transformer and LLM via Three-Word Language for Knowledge Graph Completion (2025.findings-acl)

Challenge:	Recent controllable zero-shot text-to-speech systems can synthesize speech for unseen speakers from a short reference audio clip, but they also inherit the speaking style present in the reference.
Approach:	They propose a framework that enables continuous and reference-relative style control in zero-shot text-to-speech systems by combining style-specific LoRAs with Orthogonal LoRA Fusion.
Outcome:	The proposed framework reduces the model's dependence on reference style while preserving text fidelity while maintaining intelligibility and speaker timbre.

Copied to clipboard

Kangyang Luo, Yuzhuo Bai, Cheng Gao, Shuzheng Si, Zhu Liu, Yingli Shen, Zhitong Wang, Cunliang Kong, Wenhao Li, Yufei Huang, Ye Tian, Xuantang Xiong, Lei Han, Maosong Sun

Reducing Token Redundancy in LVLMs: A Systematic Review of Token Pruning Methods (2026.acl-long)

Challenge:	Existing knowledge graphs lack the ability to integrate structural information into LLMs and output predictions deterministically.
Approach:	They propose a method which encodes structural information of KGs and merges it with LLMs to enhance KGC performance.
Outcome:	The proposed method improves the performance of KG Completion datasets on KGs by integrating structural information with LLMs.

Copied to clipboard

Hanzhang Yuan, Mengxuan Hu, Wenhao Zhang, Tianlong Wang, Zhongliang Zhou, Jiasen Lu, Sheng Li

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

Challenge:	Large Vision-Language Models (LVLMs) excel at visual understanding but face severe computational bottlenecks when processing high-resolution images and long videos due to massive visual token counts.
Approach:	They propose a taxonomy categorizing methods into vision-side, LLM-side and hybrid paradigms and analyze token selection mechanisms and pruning strategy.
Outcome:	The proposed method selectively removes less informative tokens while maintaining performance.

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

ImCoref-CeS: An Improved Lightweight Pipeline for Coreference Resolution with LLM-based Checker-Splitter Refinement (2026.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

Kangyang Luo, Yuzhuo Bai, Shuzheng Si, Cheng Gao, Zhitong Wang, Yingli Shen, Wenhao Li, Zhu Liu, Yufeng Han, Jiayi Wu, Cunliang Kong, Maosong Sun

From Scaffolding to Assimilation: Progressive Structural Internalization for Format-Constrained Creative Text Generation (2026.findings-acl)

Challenge:	Existing supervised neural methods for coreference resolution are underexplored . current methods rely on small language models, but their potential is underexploited .
Approach:	They propose a framework that integrates an enhanced supervised model with LLM-based reasoning.
Outcome:	The proposed method surpasses existing state-of-the-art methods in coreference resolution.

Copied to clipboard

Wenhao Li, Yuwei Yang, Xiaoqing Wu, Yufeng Han, Cunliang Kong, Yuzhuo Bai, Xin Cong, Maosong Sun

Text Generation

Challenge:	Existing paradigms rely on unreliable prompting or rigid constrained decoding strategies to achieve aesthetic unity.
Approach:	They propose a framework to embed external constraints into the model’s intrinsic intuition and use it to generate open-ended creative texts.
Outcome:	The proposed framework surpasses baselines in both strict constraint adherence and literary aesthetics.

MIO: A Foundation Model on Multimodal Tokens (2025.emnlp-main)

Copied to clipboard

Zekun Moore Wang, King Zhu, Chunpu Xu, Wangchunshu Zhou, Jiaheng Liu, Yibo Zhang, Jessie Wang, Ning Shi, Siyu Li, Yizhi Li, Haoran Que, Zhaoxiang Zhang, Yuanxing Zhang, Ge Zhang, Ke Xu, Jie Fu, Wenhao Huang

Debiasing Generative Named Entity Recognition by Calibrating Sequence Likelihood (2023.acl-short)

Challenge:	Existing models lack multimodal understanding capabilities, resulting in closed-source model that does not support multimodal interleaved sequences.
Approach:	They propose a foundation model built on multimodal tokens capable of understanding and generating speech, text, images, and videos in an end-to-end, autoregressive manner.
Outcome:	The proposed model is able to understand speech, text, images, and videos in an end-to-end, autoregressive manner.

Copied to clipboard

Yu Xia, Yongwei Zhao, Wenhao Wu, Sujian Li

Robustness in NLP

Challenge:	Existing approaches to recognize flat, overlapped and discontinuous entities uniformly have been used for Named Entity Recognition.
Approach:	They propose a reranking-based approach that redistributes the likelihood among candidate sequences depending on their performance via a contrastive loss.
Outcome:	The proposed method boosts baseline and yields competitive or better results compared with the state-of-the-art methods on 8 widely-used datasets for Named Entity Recognition.

Watch Every Step! LLM Agent Learning via Iterative Step-level Process Refinement (2024.emnlp-main)

Copied to clipboard

Weimin Xiong, Yifan Song, Xiutian Zhao, Wenhao Wu, Xun Wang, Ke Wang, Cheng Li, Wei Peng, Sujian Li

Training Long-Context LLMs Efficiently via Chunk-wise Optimization (2025.findings-acl)

Challenge:	Recent approaches to enhance agent performance focus on outcome rewards, which may lead to errors or suboptimal actions due to the absence of process supervision signals.
Approach:	They propose a step-level framework that provides detailed step-by-step guidance to enhance agent training by using Monte Carlo methods.
Outcome:	The proposed framework outperforms strong baselines on three tasks and shows that it is effective in augmenting efficiency and its applicability to diverse models.

Copied to clipboard

Wenhao Li, Yuxin Zhang, Gen Luo, Daohai Yu, Rongrong Ji

Multilingual Machine Translation with Large Language Models: Empirical Results and Analysis (2024.findings-naacl)

Challenge:	Recent advances in long-context large language models have demonstrated superior retrieval quality compared to retrievalaugmented generation (RAG) approaches.
Approach:	They propose a memory-efficient training paradigm that partitions lengthy inputs into manageable chunks.
Outcome:	The proposed model expands maximum sequence length from 1K to 16K tokens on a single RTX 3090 GPU, while SpaCO achieves accelerated training speed.

Copied to clipboard

Wenhao Zhu, Hongyi Liu, Qingxiu Dong, Jingjing Xu, Shujian Huang, Lingpeng Kong, Jiajun Chen, Lei Li

Multilinguality

Challenge:	Existing studies show that large language models (LLMs) can handle multilingual machine translation (MMT) However, the multilingual translation ability of LLMs remains under-explored.
Approach:	They evaluate eight popular LLMs including ChatGPT and GPT-4 to determine their performance in multilingual machine translation.
Outcome:	The proposed model can generate moderate translation even on zero-resource languages and cross-lingual exemplars can provide better task guidance for low-resourced translation than exemplar in the same language pairs.

Recurrence Boosts Diversity! Revisiting Recurrent Latent Variable in Transformer-Based Variational AutoEncoder for Diverse Text Generation (2022.findings-emnlp)

Copied to clipboard

Jinyi Hu, Xiaoyuan Yi, Wenhao Li, Maosong Sun, Xing Xie

Enhancing Zero-Shot Chain-of-Thought Reasoning in Large Language Models through Logic (2024.lrec-main)

Challenge:	Variational Auto-Encoder (VAE) has been widely adopted in text generation due to its ability to learn flexible representations.
Approach:	They propose a Transformer-based recurrent VAE structure that imposes recurrence on segment-wise latent variables with arbitrarily separated text segments and constructs the posterior distribution with residual parameterization.
Outcome:	The proposed structure can deduce a non-zero lower bound of the KL term and enhance the entanglement of each segment and preceding latent variables, providing a theoretical guarantee of generation diversity.

Copied to clipboard

Xufeng Zhao, Mengdi Li, Wenhao Lu, Cornelius Weber, Jae Hee Lee, Kun Chu, Stefan Wermter

UPLex: Fine-Grained Personality Control in Large Language Models via Unsupervised Lexical Modulation (2025.findings-emnlp)

Challenge:	Experimental evaluations of large language models demonstrate the efficacy of enhanced reasoning by logic.
Approach:	They propose a framework that uses symbolic logic to verify and rectify reasoning steps by steps.
Outcome:	The proposed framework improves the zero-shot chain-of-thought reasoning ability of large language models by verifying and rectifying the reasoning steps step by step.

Copied to clipboard

Tianlong Li, Wenhao Liu, Muling Wu, Shihan Dou, Zhenghua Wang, Changze Lv, Xiaohua Wang, Xiaoqing Zheng, Xuanjing Huang

COIG-P: A High-Quality and Large-Scale Chinese Preference Dataset for Alignment with Human Values (2026.findings-eacl)

Challenge:	Personality is a crucial factor that shapes human communication patterns, thereby regulating the personalities of large language models (LLMs).
Approach:	They propose a method that uses an Unsupervisedly-Built Personalized Lexicon (UPL) during the decoding phase to manipulate LLM’s personality traits.
Outcome:	The proposed method can modulate the personality expression of large language models by dynamically altering their predicted probability of upcoming words in a pluggable fashion.

Copied to clipboard

Siwei Wu, JinCheng Ren, Xeron Du, Shuyue Guo, Xingwei Qu, Yiming Liang, Jie Liu, Yunwen Li, Tyler Loakman, Tianyu Zheng, Boyu Feng, Huaqing Yuan, Zili Wang, Jiaheng Liu, Wenhao Huang, Chenglin Cai, Haoran Que, Jian Yang, Yuelin Bai, Zekun Moore Wang, Zhouliang Yu, Qunshu Lin, Ding Pan, Yuchen Eleanor Jiang, Tiannan Wang, Wangchunshu Zhou, Shenzhi Wang, Xingyuan Bu, Minghao Liu, Guoyin Wang, Ge Zhang, Chenghua Lin

Stylistic Chinese Poetry Generation via Unsupervised Style Disentanglement (D18-1)

Challenge:	Existing Chinese preference datasets suffer from limited scale, restricted domain coverage, and insufficiently rigorous data validation.
Approach:	They propose an LLM-based data annotation pipeline with no human intervention to annotate Chinese preference datasets.
Outcome:	The proposed pipeline outperforms existing Chinese preference datasets on AlignBench and Chinese Reward Benchmark.

Copied to clipboard

Cheng Yang, Maosong Sun, Xiaoyuan Yi, Wenhao Li

Low-Resource NLP

Challenge:	Automatic Chinese poetry generation is one of the first attempts towards computer writing.
Approach:	They propose a model which requires no supervised style labeling to generate stylistic poems . they incorporate mutual information, a concept in information theory, into modeling .
Outcome:	The proposed model generates stylistic poems without losing fluency and coherency . it is based on mutual information, a concept in information theory .

Express What You See: Can Multimodal LLMs Decode Visual Ciphers with Intuitive Semiosis Comprehension? (2025.findings-acl)

Copied to clipboard

Jiayi Kuang, Yinghui Li, Chen Wang, Haohao Luo, Ying Shen, Wenhao Jiang

Measure Twice, Click Once: Co-evolving Proposer and Visual Critic via Reinforcement Learning for GUI Grounding (2026.acl-long)

Challenge:	Traditional VQA benchmarks encounter a modality gap and over-reliance on language priors, whereas human cognition excels at intuitive semiosis, associating abstract visual symbols to linguistic semantics.
Approach:	They propose a task of generating abstract linguistics from emoji sequence images, where such reasoning underpins critical applications in cryptography.
Outcome:	The proposed model can generate abstract linguistics from emoji sequence images, challenging MLLMs’ reasoning of decoding complex semantics of visual ciphers.

Copied to clipboard

Wenkai Wang, Xiyun Li, Hongcan Guo, Wenhao Yu, Tianqing Fang, Haitao Mi, Dong Yu, Shengyu Zhang

Advancing Parameter Efficiency in Fine-tuning via Representation Editing (2024.acl-long)

Challenge:	Graphical User Interface (GUI) grounding requires mapping natural language instructions to precise pixel coordinates due to visually homogeneous elements and dense layouts.
Approach:	They propose to replace static consistency strategies with a learnable selection mechanism that selects the optimal target by critiquing its own proposals rendered on the screenshot.
Outcome:	The proposed model significantly improves both grounding and critiquing capabilities over 6 benchmarks.

Copied to clipboard

Muling Wu, Wenhao Liu, Xiaohua Wang, Tianlong Li, Changze Lv, Zixuan Ling, Zhu JianHao, Cenyuan Zhang, Xiaoqing Zheng, Xuanjing Huang

SA-CLIP: Language Guided Image Spatial and Action Feature Learning (2025.findings-emnlp)

Challenge:	Parameter Efficient Fine-Tuning (PEFT) has gained significant attention for its ability to achieve competitive results while updating only a small subset of trainable parameters.
Approach:	They propose a new approach to fine-tuning neural models that scales and biases the representation produced at each layer.
Outcome:	The proposed approach reduces the number of trainable parameters by a factor of 25,700 compared to full parameter fine-tuning and by . 32 compared with LoRA.

Copied to clipboard

Guanlin Li, Wenhao Shao, Praboda Rajapaksha, Noel Crespi

SciMMIR: Benchmarking Scientific Multi-modal Information Retrieval (2024.findings-acl)

Challenge:	Contrastive language-image pretraining models struggle with real-world downstream tasks such as road traffic anomaly detection due to inability to effectively capture spatial and action relationships between objects within images.
Approach:	They compile and curate a dataset and train a Spatial and Action relationship aware CLIP model.
Outcome:	The proposed model performs well on the traffic anomaly detection task .

Copied to clipboard

Siwei Wu, Yizhi Li, Kang Zhu, Ge Zhang, Yiming Liang, Kaijing Ma, Chenghao Xiao, Haoran Zhang, Bohao Yang, Wenhu Chen, Wenhao Huang, Noura Al Moubayed, Jie Fu, Chenghua Lin

BASS: Boosting Abstractive Summarization with Unified Semantic Graph (2021.acl-long)

Challenge:	Multi-modal information retrieval (MMIR) is a rapidly evolving field . current benchmarks for image-text pairings overlook the scientific domain .
Approach:	They develop a scientific domain-specific MMIR benchmark to evaluate image-text pairings using open-access research paper corpora.
Outcome:	The proposed benchmarks are based on 530K image-text pairs extracted from scientific documents with detailed captions.

Copied to clipboard

Wenhao Wu, Wei Li, Xinyan Xiao, Jiachen Liu, Ziqiang Cao, Sujian Li, Hua Wu, Haifeng Wang

Understanding ME? Multimodal Evaluation for Fine-grained Visual Commonsense (2022.emnlp-main)

Challenge:	Abstractive summarization for long-document or multi-document remains challenging for Seq2Seq as it does not analyze long-distance relations in text.
Approach:	They propose a framework for Boosting Abstractive Summarization based on a unified Semantic graph which aggregates co-referent phrases distributing across a long range of context and conveys rich relations between phrases.
Outcome:	The proposed framework improves document representation and summary generation process by leveraging the graph structure.

Copied to clipboard

Zhecan Wang, Haoxuan You, Yicheng He, Wenhao Li, Kai-Wei Chang, Shih-Fu Chang

MMRA: A Benchmark for Evaluating Multi-Granularity and Multi-Image Relational Association Capabilities in Large Visual Language Models (2026.findings-eacl)

Challenge:	Existing models that understand image and text but also cross-reference in-between are lacking in evaluation data resources.
Approach:	They propose a multimodal evaluation pipeline to automatically generate question-answer pairs to test models’ understanding of the visual scene, text, and related knowledge.
Outcome:	The proposed model can answer the highly semantic VCR question correctly but fails to answer related visual question (Q2), textual question (q3), and background knowledge question ( Q4) as shallow mappings with language priors and unbalanced utilization of information between modalities.

Copied to clipboard

Siwei Wu, King Zhu, Yu Bai, Yiming Liang, Yizhi Li, Haoning Wu, Jiaheng Liu, Ruibo Liu, Xingwei Qu, Xuxin Cheng, Ge Zhang, Wenhao Huang, Chenghua Lin

Ro-SLM: Onboard Small Language Models for Robot Task Planning and Operation Code Generation (2026.findings-acl)

Challenge:	Current multimodal benchmarks focus on facts within individual images, but neglect associative relations among multiple images.
Approach:	They propose a multi-image relational association task and a MMRA benchmark to evaluate LVLMs.
Outcome:	The proposed benchmarks show that entity-level multi-image perception tasks pose greater challenges than image-level tasks.

Copied to clipboard

Wenhao Wang, Yanyan Li, Long Jiao, Jiawei Yuan

Jiuge: A Human-Machine Collaborative Chinese Classical Poetry Generation System (P19-3)

Challenge:	Recent advances in large language models (LLMs) provide robots with contextual reasoning abilities to comprehend human instructions.
Approach:	They propose a framework that enables reliable SLM-driven robot operation by distilling LLMs’ knowledge and reasoning.
Outcome:	The proposed framework enables reliable SLM-driven robot operation by distilling LLMs’ knowledge and reasoning.

Copied to clipboard

Guo Zhipeng, Xiaoyuan Yi, Maosong Sun, Wenhao Li, Cheng Yang, Jiannan Liang, Huimin Chen, Yuhui Zhang, Ruoyu Li

Text Generation

Challenge:	Existing systems for automatic poetry generation are model-oriented, resulting in poor user participation.
Approach:	They propose a human-machine collaborative Chinese classical poetry generation system called Jiuge . Jiuge allows users to revise unsatisfied parts of a generated poem draft repeatedly .
Outcome:	The proposed system allows users to revise unsatisfied parts of a generated poem draft repeatedly.

Adaptive Ordered Information Extraction with Deep Reinforcement Learning (2023.findings-acl)

Copied to clipboard

Wenhao Huang, Jiaqing Liang, Zhixu Li, Yanghua Xiao, Chuanjun Ji

Revisiting Jailbreaking for Large Language Models: A Representation Engineering Perspective (2025.coling-main)

Challenge:	Existing methods for information extraction follow a fixed extraction order for complex tasks with multiple elements to be extracted in one instance.
Approach:	They propose an adaptive ordered IE paradigm to find optimal element extraction order for different instances and a reinforcement learning framework to generate optimal order dynamically.
Outcome:	The proposed method beats existing methods and improves on several public datasets.

Copied to clipboard

Tianlong Li, Zhenghua Wang, Wenhao Liu, Muling Wu, Shihan Dou, Changze Lv, Xiaohua Wang, Xiaoqing Zheng, Xuanjing Huang

Fuse It More Deeply! A Variational Transformer with Layer-Wise Latent Variable Inference for Text Generation (2022.naacl-main)

Challenge:	Recent surge in jailbreaking attacks has revealed significant vulnerabilities in Large Language Models (LLMs) however, limited research into the underlying mechanisms that make LLMs vulnerable to such attacks has been conducted.
Approach:	They propose that LLMs' self-safeguarding capability is linked to specific activity patterns within their representation space.
Outcome:	The proposed models can be detected with a few pairs of contrastive queries, and the robustness can be manipulated by weakening or strengthening these patterns.

Copied to clipboard

Jinyi Hu, Xiaoyuan Yi, Wenhao Li, Maosong Sun, Xing Xie

Dataset Bias Mitigation in Multiple-Choice Visual Question Answering and Beyond (2023.findings-emnlp)

Challenge:	Variational Auto-Encoders are often used for text generation tasks due to the sequential nature of the text.
Approach:	They propose a variational Transformer framework that learns a series of layer-wise latent variables with each inferred from those of lower layers and tightly coupled with the hidden states by low-rank tensor product.
Outcome:	The proposed framework can learn latent variables from lower layers and incorporate more information.

Copied to clipboard

Zhecan Wang, Long Chen, Haoxuan You, Keyang Xu, Yicheng He, Wenhao Li, Noel Codella, Kai-Wei Chang, Shih-Fu Chang

Composing Elementary Discourse Units in Abstractive Summarization (2020.acl-main)

Challenge:	Existing studies have examined dataset biases in VQA benchmarks with short-phrase answers Multiple-choice Question with the LONG Answers (VCR, VLEP, etc.)
Approach:	They propose to use Adversarial Data Synthesis (ADS) to generate synthetic training and debiased evaluation data and introduce Intra-sample Counterfactual Training (ICT) to assist models in utilizing synthesized training data.
Outcome:	The proposed approach improves model performance even in domain-shifted scenarios.

Copied to clipboard

Zhenwen Li, Wenhao Wu, Sujian Li

Summarization

Challenge:	Abstractive summarization uses a single document sentence to generate a summary, but this can cause performance degradation.
Approach:	They propose to use elementary discourse unit (EDU) as the summarization unit to extract and group informative EDUs and then an EDU fusion model to fuse the EDU in each group into one sentence.
Outcome:	The proposed model can be used to combine informative EDUs into one sentence and reward selection actions.

LongEmbed: Extending Embedding Models for Long Context Retrieval (2024.emnlp-main)

Copied to clipboard

Dawei Zhu, Liang Wang, Nan Yang, Yifan Song, Wenhao Wu, Furu Wei, Sujian Li

Representation Learning

Challenge:	Existing embedding models support only 512 input tokens, hindering their application in scenarios requiring long inputs.
Approach:	They evaluate the performance of existing embedding models by using a new benchmark and a training-free context window extension strategy.
Outcome:	The proposed model extends the input window of existing models by several folds.

BenchMAX: A Comprehensive Multilingual Evaluation Suite for Large Language Models (2025.findings-emnlp)

Copied to clipboard

Xu Huang, Wenhao Zhu, Hanxu Hu, Conghui He, Lei Li, Shujian Huang, Fei Yuan

CoUDA: Coherence Evaluation via Unified Data Augmentation (2024.naacl-long)

Challenge:	Existing multilingual benchmarks focus primarily on language understanding tasks.
Approach:	They develop a multi-way multilingual benchmark that measures critical capabilities of large language models across languages.
Outcome:	Extensive experiments on BenchMAX reveal uneven utilization of core capabilities across languages, emphasizing the performance gaps that scaling model size alone does not resolve.

Copied to clipboard

Dawei Zhu, Wenhao Wu, Yifan Song, Fangwei Zhu, Ziqiang Cao, Sujian Li

CoTJudger: A Graph-Driven Framework for Automatic Evaluation of Chain-of-Thought Efficiency and Redundancy in LRMs (2026.findings-acl)

Challenge:	Existing data augmentations for coherence evaluation rely on heuristic rules and lack designing criteria.
Approach:	They propose a data augmentation framework that breaks down coherence into global and local aspects and designs augmentation strategies for both aspects.
Outcome:	The proposed framework surpasses recent models in scoring and ranking tasks with 233M parameters.

Copied to clipboard

Siyi Li, Jiajun Shi, Shiwen Ni, Ge Zhang, Shuaimin Li, Shijian Wang, Zhoufutu Wen, Yizhi LI, Hamid Alinejad-Rokny, Jiaheng Liu, Min Yang, Wenhao Huang

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

Challenge:	Existing evaluations emphasize final accuracy or coarse token counts, and lack automated tools to separate essential logic from structural redundancy.
Approach:	They propose a graph-driven framework that quantifies reasoning efficiency by converting free-form CoTs into directed dependency graphs and extracting the Shortest Effective Path needed to reach a correct solution.
Outcome:	Evaluating 21 LRMs, the proposed framework quantifies reasoning efficiency by converting free-form CoTs into directed dependency graphs and extracting the Shortest Effective Path (SEP) needed to reach a correct solution.

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

Chain-of-Note: Enhancing Robustness in Retrieval-Augmented Language Models (2024.emnlp-main)

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

Wenhao Yu, Hongming Zhang, Xiaoman Pan, Peixin Cao, Kaixin Ma, Jian Li, Hongwei Wang, Dong Yu

RAISE: Reinforced Adaptive Instruction Selection For Large Language Models (2025.findings-emnlp)

Challenge:	Standard RALMs often neglect their intrinsic knowledge due to the interference from retrieved information.
Approach:	They propose a new approach to improve robustness of RALMs by generating sequential reading notes for each retrieved document.
Outcome:	The proposed approach outperforms standard RALMs on four open-domain QA benchmarks.

Copied to clipboard

Qingsong Lv, Yangning Li, Zihua Lan, Zishan Xu, Jiwei Tang, Tingwei Lu, Yinghui Li, Wenhao Jiang, Hong-Gee Kim, Hai-Tao Zheng, Philip S. Yu

Promoting Data and Model Privacy in Federated Learning through Quantized LoRA (2024.findings-emnlp)

Challenge:	Existing selection methods rely on static, heuristic quality scores and are executed only once before training.
Approach:	They propose a dynamic selection framework that integrates selection into every training step.
Outcome:	The proposed framework integrates selection into every training step.

Copied to clipboard

Zhu JianHao, Changze Lv, Xiaohua Wang, Muling Wu, Wenhao Liu, Tianlong Li, Zixuan Ling, Cenyuan Zhang, Xiaoqing Zheng, Xuanjing Huang

DynaCode: A Dynamic Complexity-Aware Code Benchmark for Evaluating Large Language Models in Code Generation (2025.findings-acl)

Challenge:	Existing federated learning frameworks require substantial data and computational resources to develop large language models.
Approach:	They propose a method that distributes a quantized version of the model’s parameters during training and combine it with a popular fine-tuning method to significantly reduce communication costs.
Outcome:	The proposed method enables accurate estimations for parameter updates while preventing clients from accessing a model whose performance is comparable to the centrally hosted one.

Copied to clipboard

Wenhao Hu, Jinhao Duan, Chunchen Wei, Li Zhang, Yue Zhang, Kaidi Xu

Chain-of-Thought Matters: Improving Long-Context Language Models with Reasoning Path Supervision (2025.findings-emnlp)

Challenge:	Existing code benchmarks for large language models remain static, resulting in data contamination and unreliable evaluation results.
Approach:	They propose a dynamic, complexity-aware benchmark that overcomes the limitations of static datasets and provides a memorization-advantaged benchmark.
Outcome:	DynaCode generates 189 million unique nested code problems across 4 units of code complexity and 16 types of call graphs.

Copied to clipboard

Dawei Zhu, Xiyu Wei, Guangxiang Zhao, Wenhao Wu, Haosheng Zou, Junfeng Ran, null XWang, Lin Sun, Xiangzheng Zhang, Sujian Li