Papers by Dongsheng Li

GCML: Gradient Coherence Guided Meta-Learning for Cross-Domain Emerging Topic Rumor Detection (2025.emnlp-main)

Challenge:	Existing methods for LGT detection assume that it is a single homogeneous distribution.
Approach:	They propose a framework for LGT detection based on density-aware manifold learning and hybrid Mahalanobis energy.
Outcome:	The proposed framework outperforms baselines in detecting LLM-generated text (LGT) it is based on density-aware manifold learning and hybrid Mahalanobis energy .

Copied to clipboard

Zejiang He, Jingyuan Huang, Menglong Lu, Zhen Huang, Shanshan Liu, Zhiliang Tian, Dongsheng Li

Attention-Guided Answer Distillation for Machine Reading Comprehension (D18-1)

Challenge:	Existing domain adaptation rumor detection methods ignore the data generalization differences and rely on a large amount of unlabeled target domain samples to achieve domain adaptation.
Approach:	They propose a Gradient Coherence guided Meta-Learning approach for emerging topics rumor detection that selectively learns more "generalizable" tasks that are more beneficial in adapting to the target domain.
Outcome:	The proposed method outperforms baselines on real-world datasets and significantly outperformed traditional methods on the in-domain condition.

Copied to clipboard

Minghao Hu, Yuxing Peng, Furu Wei, Zhen Huang, Dongsheng Li, Nan Yang, Ming Zhou

Context-aware Watermark with Semantic Balanced Green-red Lists for Large Language Models (2024.emnlp-main)

Challenge:	Existing approaches to reading comprehension systems are vulnerable to adversarial attacks.
Approach:	They propose to use knowledge distillation to transfer knowledge from an ensemble to a single model.
Outcome:	The proposed methods outperform the teacher on adversarial datasets and NarrativeQA benchmarks.

Copied to clipboard

Yuxuan Guo, Zhiliang Tian, Yiping Song, Tianlun Liu, Liang Ding, Dongsheng Li

Robustness in NLP

Challenge:	Recent research suggests that watermarking methods cause degradation of text quality due to semantic disparities between the watermarked text and the unwatermarked text.
Approach:	They propose a semantic-aware watermark method that generates a watermark key considering contexts to split a green/red list for watermark injection.
Outcome:	The proposed method reduces performance drop due to adding bias on green lists . it also allows green lists to cover almost all semantics .

MinerU2.5: A Decoupled Vision-Language Model for Efficient High-Resolution Document Parsing (2026.acl-industry)

Copied to clipboard

Junbo Niu, Zheng Liu, Zhuangcheng Gu, Bin Wang, Linke Ouyang, Zhiyuan Zhao, Tao Chu, Tianyao He, Fan Wu, Qintong Zhang, Zhenjiang Jin, Guang Liang, Rui Zhang, Wenzheng Zhang, Yuan Qu, Zhifei Ren, Yuefeng Sun, Zirui Tang, Boyu Niu, Yuanhong Zheng, Dongsheng Ma, Ziyang Miao, Hejun Dong, Siyi Qian, Junyuan Zhang, Fangdong Wang, Jingzhou Chen, Xiaomeng Zhao, Liqun Wei, Wei Li, Shasha Wang, RuiLiang Xu, Yuanyuan Cao, Lu Chen, Qianqian Wu, Huaiyu Gu, Lindong Lu, Dechen Lin, null Shenguanlin, Xuanhe Zhou, Linfeng Zhang, Yuhang Zang, Xiaoyi Dong, Jiaqi Wang, Bo Zhang, Lei Bai, Pei Chu, Weijia Li, Jiang Wu, Lijun Wu, Zhenxiang Li, Guangyu Wang, Zhongying Tu, Chao Xu, Kai Chen, Bowen Zhou, Dahua Lin, Wentao Zhang, Conghui He

Green & Sustainable NLP

Challenge:	Document images are characterized by higher resolutions, denser content, and more complex structural layouts.
Approach:	They propose a 1.2B-parameter document parsing vision-language model that decouples layout analysis from local content recognition.
Outcome:	The proposed model surpasses general-purpose and domain-specific models on multiple benchmarks while maintaining significantly lower computational overhead.

AdaMARP: An Adaptive Multi-Agent Interaction Framework for General Immersive Role-Playing (2026.findings-acl)

Copied to clipboard

Zhenhua Xu, Dongsheng Chen, Shuo Wang, Jian Li, Chengjie Wang, Meng Han, Yabiao Wang

Challenge:	Existing LLMs lack immersion and adaptability, resulting in limited character orchestration and on-the-fly character introduction.
Approach:	They propose an LLM-based framework that allows actors to interact with users in an ongoing narrative.
Outcome:	The proposed framework outperforms commercial LLMs in character consistency, environment grounding, and narrative coherence.

Sparrow: Text-Anchored Window Attention with Visual-Semantic Glimpsing for Speculative Decoding in Video LLMs (2026.acl-long)

Copied to clipboard

Libo Zhang, Zhaoning Zhang, null Hongwanyang, Peng Qiao, Dongsheng Li

Visual Data in NLP

Challenge:	Existing speculative decoding models face performance collapse due to key-value cache explosion and context window mismatches.
Approach:	They propose a framework that offloads visual computation to the target model by using hidden state reuse.
Outcome:	The proposed framework achieves an average speedup of 2.82x even with 25k visual tokens .

Social Bot-Aware Graph Neural Network for Early Rumor Detection (2022.coling-1)

Copied to clipboard

Zhen Huang, Zhilong Lv, Xiaoyun Han, Binyang Li, Menglong Lu, Dongsheng Li

Structured Data in NLP

Challenge:	Existing models do not distinguish genuine users from social bots, and their failure in identifying rumors timely.
Approach:	They propose to account for social bots’ behavior and construct a Social Bot-Aware Graph Neural Network to model early propagation of posts and then use it to detect rumors.
Outcome:	The proposed method achieves significant improvements over baselines and identifies rumors within 3 hours while maintaining more than 90% accuracy.

POMP: Probability-driven Meta-graph Prompter for LLMs in Low-resource Unsupervised Neural Machine Translation (2024.acl-long)

Copied to clipboard

Shilong Pan, Zhiliang Tian, Liang Ding, Haoqi Zheng, Zhen Huang, Zhihua Wen, Dongsheng Li

Low-Resource NLP

Challenge:	Low-resource languages (LRLs) face challenges in supervised neural machine translation due to limited parallel data.
Approach:	They propose a method that uses a dynamic graph to organize auxiliary languages in prompts to improve LRL translations.
Outcome:	The proposed method improves translation accuracy in low-resource languages (LRLs) using auxiliary language pairs and synthetic pseudo-parallel data.

Mitigate Position Bias in LLMs via Scaling a Single Hidden States Channel (2025.findings-acl)

Copied to clipboard

Yijiong Yu, Huiqiang Jiang, Xufang Luo, Qianhui Wu, Chin-Yew Lin, Dongsheng Li, Yuqing Yang, Yongfeng Huang, Lili Qiu

Challenge:	Long-context language models exhibit position bias, also known as "lost in the middle" research shows that even long-contemporary LLMs fail to utilize all context information effectively .
Approach:	They propose a method to mitigate position bias by scaling positional hidden states . they propose to use a channel of hidden states to modify positional Hidden states a LCLM's positional bias .
Outcome:	The proposed method can improve performance by 15.2% in a "lost in the middle" benchmark.

LLM-based Rumor Detection via Influence Guided Sample Selection and Game-based Perspective Analysis (2025.acl-long)

Copied to clipboard

Zhiliang Tian, Jingyuan Huang, Zejiang He, Zhen Huang, Menglong Lu, Linbo Qiao, Songzhu Mei, Yijie Wang, Dongsheng Li

Challenge:	Existing methods for rumor detection on social media are limited by limited modeling capacity and insufficient training corpora.
Approach:	They propose an SFT-based rumor detection model with Influence guided Sample selection and Game-based multi-perspective analysis to address these issues.
Outcome:	The proposed model outperforms existing SOTA on three datasets.

EvoAgent: Towards Automatic Multi-Agent Generation via Evolutionary Algorithms (2025.naacl-long)

Copied to clipboard

Siyu Yuan, Kaitao Song, Jiangjie Chen, Xu Tan, Dongsheng Li, Deqing Yang

Towards Understanding Omission in Dialogue Summarization (2023.acl-long)

Challenge:	Existing work on extending specialized agents to multi-agent systems is dependent on human-designed frameworks, limiting the functional scope and scalability of agent systems.
Approach:	They propose a generic method to automatically extend specialized agents to multi-agent systems via evolutionary algorithm . they consider existing agent frameworks as the initial individual and apply evolutionary operators to generate multiple agents with diverse settings.
Outcome:	The proposed method can extend specialized agents to multi-agent systems . it can generate multiple agents with diverse settings, and improves performance across tasks .

Copied to clipboard

Yicheng Zou, Kaitao Song, Xu Tan, Zhongkai Fu, Qi Zhang, Dongsheng Li, Tao Gui

Dialogue Systems & Conversational Agents

Challenge:	Existing methods for dialogue summarization are far from satisfactory . omission is a major factor in affecting the quality of summarizing, but few studies have explored the problem .
Approach:	They propose a dataset that provides high-quality omission labels for dialogue summarization . they propose to use this dataset to detect omitted dialogue utterances .
Outcome:	The proposed dataset improves summarization quality by providing ground-truth omission labels . the proposed dataset and codes are publicly available .

Dovetail: A CPU/GPU Heterogeneous Speculative Decoding for LLM inference (2025.emnlp-main)

Copied to clipboard

Libo Zhang, Zhaoning Zhang, null Xubaizhou, Rui Li, Zhiliang Tian, Songzhu Mei, Dongsheng Li

Green & Sustainable NLP

Challenge:	Large language models (LLMs) are demanding more memory and computational resources . however, these devices typically feature weaker GPUs and stronger CPUs .
Approach:	They propose a lossless inference acceleration method that leverages the characteristics of heterogeneous devices and the advantages of speculative decoding.
Outcome:	The proposed method achieves speedups ranging from 1.79 to 10.1 across different devices . it uses a draft model on the GPU to perform preliminary predictions, while a target model on CPU validates these outputs .

Two-stage Generative Question Answering on Temporal Knowledge Graph Using Large Language Models (2024.findings-acl)

Copied to clipboard

Yifu Gao, Linbo Qiao, Zhigang Kan, Zhihua Wen, Yongquan He, Dongsheng Li

Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and Classification (P19-1)

Challenge:	Temporal knowledge graph question answering (TKGQA) is one of the most challenging QA tasks due to the temporal constraints hidden in questions and the answers sought from dynamic structured knowledge.
Approach:	They propose a generative temporal knowledge graph question answering framework which guides LLMs to answer temporal questions through two phases: Subgraph Retrieval and Answer Generation.
Outcome:	The proposed framework exploits LLM’s intrinsic knowledge to mine temporal constraints and structural links in the questions without extra training, thus narrowing down the subgraph search space in both temporal and structural dimensions.

Copied to clipboard

Minghao Hu, Yuxing Peng, Zhen Huang, Dongsheng Li, Yiwei Lv

Emancipating Event Extraction from the Constraints of Long-Tailed Distribution Data Utilizing Large Language Models (2024.lrec-main)

Challenge:	Existing approaches to target sentiment analysis are limited by huge search space and sentiment inconsistency.
Approach:	They propose a span-based extract-then-classify framework to detect opinion targets . they propose pipeline, joint, and collapsed models to classify polarities .
Outcome:	The proposed framework outperforms the sequence tagging baseline on three benchmark datasets.

Copied to clipboard

Zhigang Kan, Liwen Peng, Linbo Qiao, Dongsheng Li

Self-Evolution Learning for Mixup: Enhance Data Augmentation on Few-Shot Text Classification Tasks (2023.emnlp-main)

Challenge:	Existing methods for EE depend on manual annotations, which are expensive and scarce.
Approach:	They propose to transform the event extraction task into multi-turn dialogues and a novel method for generating high-quality data.
Outcome:	The proposed methods significantly improve existing models’ performance with various paradigms and structures, especially on tail types.

Copied to clipboard

Haoqi Zheng, Qihuang Zhong, Liang Ding, Zhiliang Tian, Xin Niu, Changjian Wang, Dongsheng Li, Dacheng Tao

Low-Resource NLP

Challenge:	Text classification tasks often encounter few-shot scenarios with limited labeled data, and addressing data scarcity is crucial.
Approach:	They propose a self-evolution learning (SE) based mixup approach for data augmentation in text classification which generates more adaptive and model-friendly pseudo samples for the model training.
Outcome:	The proposed approach can generate more adaptive and model-friendly pseudo samples for the model training.

MLCopilot: Unleashing the Power of Large Language Models in Solving Machine Learning Tasks (2024.eacl-long)

Copied to clipboard

Lei Zhang, Yuge Zhang, Kan Ren, Dongsheng Li, Yuqing Yang

CTTA-T: Continual Test-Time Adaptation for Text Understanding via Teacher-Student with a Domain-aware and Generalized Teacher (2026.acl-long)

Challenge:	Existing approaches to automating ML are time-consuming and difficult to understand for human developers.
Approach:	They propose a framework that leverages large language models to develop ML solutions for novel tasks.
Outcome:	The proposed framework bridges the gap between machine intelligence and human knowledge by exploiting state-of-the-art large language models.

Copied to clipboard

Tianlun Liu, Zhiliang Tian, Zhen Huang, Xingzhi Zhou, Wanlong Yu, Tianle Liu, Feng Liu, Dongsheng Li

DYNTEXT: Semantic-Aware Dynamic Text Sanitization for Privacy-Preserving LLM Inference (2025.findings-acl)

Challenge:	Existing models for text understanding fail to adapt to domain shifts in real-world applications . current models do not improve themselves as they are applied to new domains .
Approach:	They propose a continual test-time adaptation framework that adapts to evolving domains . they propose accumulating domains and a refine-then-filter framework to calibrate teacher predictions .
Outcome:	The proposed model excels in a teacher-student framework adaptable to evolving domains.

Copied to clipboard

Juhua Zhang, Zhiliang Tian, Minghang Zhu, Yiping Song, Taishu Sheng, Siyi Yang, Qiunan Du, Xinwang Liu, Minlie Huang, Dongsheng Li

DaMSTF: Domain Adversarial Learning Enhanced Meta Self-Training for Domain Adaptation (2023.acl-long)

Challenge:	Existing methods to protect privacy of sensitive data are differential privacy (DP) and DP is used to protect users from privacy leakage.
Approach:	They propose an LDP-based Dynamic Text sanitization for privacy-preserving LLM inference that dynamically constructs semantic-aware adjacency lists of sensitive tokens to sample non-sensitive tokens for perturbation.
Outcome:	The proposed model excels on three datasets.

Copied to clipboard

Menglong Lu, Zhen Huang, Yunxiang Zhao, Zhiliang Tian, Yang Liu, Dongsheng Li

MONTROSE: LLM-driven Monte Carlo Tree Search Self-Refinement for Cross-Domain Rumor Detection (2025.findings-acl)

Challenge:	Existing approaches to domain adaptation only use reliable pseudo instances, i.e., pseudo instances with high prediction confidence, to retrain the model.
Approach:	They propose a domain adversarial learning enhanced self-training framework that uses meta-learning to estimate the importance of each pseudo instance and a meta constructor to construct the meta-validation set.
Outcome:	The proposed framework reduces label noise and preserves hard examples while maintaining accuracy.

Copied to clipboard

Shanshan Liu, Menglong Lu, Zhen Huang, Zejiang He, Liu Liu, Zhigang Sun, Dongsheng Li

Information Retrieval

Challenge:	Existing feature alignment methods are susceptible to task interference during training.
Approach:	MONTROSE is a cross-domain rumor detection method that generates high-quality synthetic data for the target domain and a domain-sharpness-aware approach to train models with these synthetic data.
Outcome:	Experiments show that MONTROSE improves in cross-domain rumor detection.

GRACE: Gradient-guided Controllable Retrieval for Augmenting Attribute-based Text Generation (2023.findings-acl)

Copied to clipboard

Zhihua Wen, Zhiliang Tian, Zhen Huang, Yuxin Yang, Zexin Jian, Changjian Wang, Dongsheng Li

DiffusionNER: Boundary Diffusion for Named Entity Recognition (2023.acl-long)

Challenge:	Existing methods for controlling the generation of pre-trained language models infuse domain bias into the generation process, making it difficult to generate out-of-domain texts.
Approach:	They propose a retrieval-augmented generation framework that uses retrieval to generate fluent sentences with high attribute relevance.
Outcome:	The proposed method can generate fluent sentences with high attribute relevance while keeping domain bias out of the model.

Copied to clipboard

Yongliang Shen, Kaitao Song, Xu Tan, Dongsheng Li, Weiming Lu, Yueting Zhuang

Correlation-Aware Example Selection for In-Context Learning with Nonsymmetric Determinantal Point Processes (2025.emnlp-main)

Challenge:	Named Entity Recognition (NER) tasks are fundamental to many structured information extraction tasks.
Approach:	They propose a named entity recognition task that uses a boundary-denoising diffusion process to denoise noisy spans.
Outcome:	The proposed method achieves comparable or even better performance than previous state-of-the-art models on flat and nested datasets.

Copied to clipboard

Qiunan Du, Zhiliang Tian, Zhen Huang, Kailun Bian, Tianlun Liu, Zhaoning Zhang, Xinwang Liu, Feng Liu, Dongsheng Li

Low-Resource NLP

Challenge:	Existing studies on in-context learning (ICL) focus on the selection of individual examples and ignore correlations among examples.
Approach:	They propose a method to capture positive and negative correlations using the determinantal point process . they optimize the method via kernel decomposition-based MLE to fit a constructed pseudo-labeled dataset .
Outcome:	The proposed method outperforms baselines in ICL example selection.

Emotion Trajectory-aware Retrieval for Markov-driven Emotion Anticipation in LLM-based Emotional Support Conversation (2026.findings-acl)

Copied to clipboard

Hongyan Wu, Zhiliang Tian, Zhen Huang, Tengyue Hu, Linbo Qiao, Yifu Gao, Feng Liu, Dongsheng Li

Information Retrieval

Challenge:	Existing strategies focus on planning the next-turn dialogue strategies, while external strategy planners focus on generating empathetic responses.
Approach:	They propose a Markov-driven emotion anticipation framework with emotion trajectory-aware retrieval for LLM-based ESC, which anticipates future emotion states to guide strategy planning and achieve sustained emotional support.
Outcome:	The proposed framework can anticipate future emotions and achieve sustained emotional support on two datasets with two models.

A Multi-Type Multi-Span Network for Reading Comprehension that Requires Discrete Reasoning (D19-1)

Copied to clipboard

Minghao Hu, Yuxing Peng, Zhen Huang, Dongsheng Li

Reasoning

Challenge:	Existing models for reading comprehension and question answering do not support discrete reasoning abilities.
Approach:	They propose a reading comprehension model that uses a multi-type answer predictor and a multiple-span extraction method to produce one or multiple text spans.
Outcome:	The proposed model achieves 79.9 F1 on the DROP hidden test set, creating new state-of-the-art results.

Dynamic PMI-Guided Contrastive Decoding Reduces Hallucination in Large Language Models: A Unified Framework of Fine-Grained Input Transformations (2026.findings-acl)

Copied to clipboard

Dongsheng Chen, Yingqi Zhu, Xingyue Zhang, Wenqing Zhou, Lei Li

EASYTOOL: Enhancing LLM-based Agents with Concise Tool Instruction (2025.naacl-long)

Challenge:	Despite the remarkable generation capabilities of large language models, the issue of hallucination remains a critical challenge.
Approach:	They propose a contrastive decoding framework based on dynamic pointwise mutual information that disentangles spurious dependencies induced by context priors, lexical co-occurrences, and syntactic structures and prioritizes causal logic.
Outcome:	The proposed framework significantly improves the model’s factuality and reasoning robustness while maintaining high computational efficiency.

Copied to clipboard

Siyu Yuan, Kaitao Song, Jiangjie Chen, Xu Tan, Yongliang Shen, Kan Ren, Dongsheng Li, Deqing Yang

GRASS: Gradient-based Adaptive Layer-wise Importance Sampling for Memory-efficient Large Language Model Fine-tuning (2026.findings-acl)

Challenge:	EASYTOOL combines tools from diverse tool documentation into a single tool instruction.
Approach:	They propose a framework that transforms tool documentation into a unified tool instruction.
Outcome:	EASYTOOL combines extensive tool documentation into a concise tool instruction . it reduces token consumption and improves performance of LLM-based agents .

Copied to clipboard

Kaiyuan Tian, Linbo Qiao, Yu Tang, Gongqingjian Jiang, Baihui Liu, Yifu Gao, Xialin Su, Dongsheng Li

Empathetic and Emotionally Positive Conversation Systems with an Emotion-specific Query-Response Memory (2022.findings-emnlp)

Challenge:	Low-rank adaptation methods for large language models limit expressiveness and performance . layer-wise fine-tuning methods overlook variations in layer importance across tasks and training stages, resulting in suboptimal performance on downstream tasks.
Approach:	They propose a gradient-based adaptive layer-wise importance sampling framework that updates only a subset of parameters to reduce memory usage.
Outcome:	The proposed framework outperforms state-of-the-art methods in accuracy and memory usage.

Copied to clipboard

Zhiliang Tian, Yinliang Wang, Yiping Song, Chi Zhang, Dongkyu Lee, Yingxiu Zhao, Dongsheng Li, Nevin L. Zhang

Dialogue Systems & Conversational Agents

Challenge:	Existing emotional conversation systems output responses according to either a given emotion or the user’s emotion reflected in the input queries.
Approach:	They propose to generate empathetic responses catering to the user’s emotions while leading the conversation to be emotionally positive by abstracting the conversation corpus and extracting the different responding strategies for different users’ emotions and conversational topics into a memory.
Outcome:	The proposed model surpasses the baseline methods in appropriateness, diversity, and generating emotionally positive responses.

DPGA-TextSyn: Differentially Private Genetic Algorithm for Synthetic Text Generation (2025.findings-acl)

Copied to clipboard

Zhonghao Sun, Zhiliang Tian, Yiping Song, Yuyi Si, Juhua Zhang, Minlie Huang, Kai Lu, Zeyu Xiong, Xinwang Liu, Dongsheng Li

Ethical NLP

Challenge:	Existing methods to fine-tune large language models pose privacy risks . researchers have synthesized data with strong generation capabilities closed-source LLMs to alleviate this problem .
Approach:	They propose to combine general LLMs with genetic algorithm to produce relevant and diverse synthetic text under differential privacy constraints.
Outcome:	The proposed method significantly improves the performance of the model in downstream tasks while ensuring privacy.

Improving Factual Consistency in Abstractive Summarization with Sentence Structure Pruning (2024.lrec-main)

Copied to clipboard

Dingxin Hu, Xuanyu Zhang, Xingyue Zhang, Yiyang Li, Dongsheng Chen, Marina Litvak, Natalia Vanetik, Qing Yang, Dongliang Xu, Yanquan Zhou, Lei Li, Yuze Li, Yingqi Zhu

Alloc-MoE: Budget-Aware Expert Activation Allocation for Efficient Mixture-of-Experts Inference (2026.acl-long)

Challenge:	Abstractive summarization models suffer from factual inconsistency problem . post-editing methods focus on replacing suspicious entities, failing to modify incorrect content hidden in sentence structures.
Approach:	They propose to use sentence pruning operation to correct possible errors . they propose to apply sentence pruning operations to the syntactic dependency tree .
Outcome:	The proposed method improves factual consistency on the FRANK dataset compared with baselines . it is model-independent and can serve as the final step in ensuring factual consistentness.

Copied to clipboard

Baihui Liu, Kaiyuan Tian, Wei Wang, Zhaoning Zhang, Linbo Qiao, Dongsheng Li

Green & Sustainable NLP

Challenge:	Existing approaches that reduce expert activations lead to severe model performance degradation.
Approach:	They propose a framework that optimizes budget allocation coordinately at layer and token levels to minimize model performance degradation.
Outcome:	The proposed framework achieves 1.15 prefill and 1.34 decode speedups on DeepSeek-V2-Lite at half of the original budget.

Exploring Pre-trained Language Models for Event Extraction and Generation (P19-1)

Copied to clipboard

Sen Yang, Dawei Feng, Linbo Qiao, Zhigang Kan, Dongsheng Li

GROVE: A Retrieval-augmented Complex Story Generation Framework with A Forest of Evidence (2023.findings-emnlp)

Challenge:	Existing methods to extract event data are laborious to create and limited in size.
Approach:	They propose an event extraction model to overcome the roles overlap problem by separating the argument prediction in terms of roles.
Outcome:	The proposed method surpasses existing methods on the ACE2005 dataset and improves on the previous methods.

Copied to clipboard

Zhihua Wen, Zhiliang Tian, Wei Wu, Yuxin Yang, Yanqi Shi, Zhen Huang, Dongsheng Li

Learning Joint Structural and Temporal Contextualized Knowledge Embeddings for Temporal Knowledge Graph Completion (2023.findings-acl)

Challenge:	Existing methods for generating stories with complex plots rely on detailed prompts, which inadvertently limit the creative potential of the generated stories.
Approach:	They propose a retrieval-auGmented stoRy generation framework with a fOrest of eVidEnce to enhance stories’ complexity.
Outcome:	The proposed framework enables generating more diverse plotlines from human-written stories.

Copied to clipboard

Yifu Gao, Yongquan He, Zhigang Kan, Yi Han, Linbo Qiao, Dongsheng Li

Improving Large Language Models in Event Relation Logical Prediction (2024.acl-long)

Challenge:	Existing methods that incorporate time information into static knowledge graph embedding ignore the contextual nature of the TKG structure.
Approach:	They propose a method that employs pre-trained language models to learn joint Structural and Temporal Contextualized Knowledge Embeddings.
Outcome:	The proposed method is superior to existing methods that ignore the contextual nature of the TKG structure.

Copied to clipboard

Meiqi Chen, Yubo Ma, Kaitao Song, Yixin Cao, Yan Zhang, Dongsheng Li

KC-GenRe: A Knowledge-constrained Generative Re-ranking Method Based on Large Language Models for Knowledge Graph Completion (2024.lrec-main)

Challenge:	Event relation extraction tasks require rigorous logical reasoning and semantic comprehension, a challenge for narrative understanding and reasoning.
Approach:	They propose three approaches to endow LLMs with event relation logic to generate more coherent answers across different scenarios.
Outcome:	The proposed approach improves on a set of ERE tasks and provides insights for future work.

Copied to clipboard

Yilin Wang, Minghao Hu, Zhen Huang, Dongsheng Li, Dong Yang, Xicheng Lu

Hierarchical Safety Realignment: Lightweight Restoration of Safety in Pruned Large Vision-Language Models (2025.findings-acl)

Challenge:	Knowledge graph completion (KGC) is a critical task to predict missing facts among entities.
Approach:	They propose a knowledge-constrained generative re-ranking method based on generative large language models for KGC that can predict missing facts among entities.
Outcome:	The proposed method achieves state-of-the-art performance on four datasets and 9.0% and 11.1% compared to the previous methods.

Copied to clipboard

Yue Li, Xin Yi, Dongsheng Shi, Gerard De Melo, Xiaoling Wang, Linlin Wang

Diversity and Consistency: Exploring Visual Question-Answer Pair Generation (2021.findings-emnlp)

Challenge:	Recent work has shown that pruning can reduce model performance, but it can also lead to degradation in safety performance.
Approach:	They propose a hierarchical safety realignment approach to prune large vision-Language Models . they quantify contribution of each attention head to safety and restore neurons .
Outcome:	The proposed approach achieves significant safety improvements in LVLMs pruned post pruning.

Copied to clipboard

Sen Yang, Qingyu Zhou, Dawei Feng, Yang Liu, Chao Li, Yunbo Cao, Dongsheng Li

Multimodality

Challenge:	Existing tasks to generate question-answer pairs from visual images are under-explored.
Approach:	They propose a task that targets question-answer pair generation from visual images.
Outcome:	The proposed model can generate diverse or consistent QAPs on two benchmarks.

Improving Few-Shot Relation Classification by Prototypical Representation Learning with Definition Text (2022.findings-naacl)

Copied to clipboard

Li Zhenzhen, Yuyang Zhang, Jian-Yun Nie, Dongsheng Li

Retrieve, Read, Rerank: Towards End-to-End Multi-Document Reading Comprehension (P19-1)

Challenge:	Existing approaches to few-shot relation classification have limited labeled examples . a prototype encoder from definition and an instance is needed to learn relation instance classification .
Approach:	They propose to learn a prototype encoder from relation definition in a way that is useful for relation instance classification.
Outcome:	The proposed encoder outperforms state-of-the-art methods on several datasets.

Copied to clipboard

Minghao Hu, Yuxing Peng, Zhen Huang, Dongsheng Li

Reasoning

Challenge:	Existing approaches to answer reading comprehension tasks are inefficient since the input is re-encoded within each module.
Approach:	They propose a unified question answering model that combines context retrieving, reading comprehension, and answer reranking to predict the final answer.
Outcome:	The proposed model outperforms the baseline model and achieves state-of-the-art results on two versions of TriviaQA and two variants of SQuAD.

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

Copied to clipboard

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

LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compression (2024.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

Huiqiang Jiang, Qianhui Wu, Xufang Luo, Dongsheng Li, Chin-Yew Lin, Yuqing Yang, Lili Qiu