Papers by Zhang Chi

ET5: A Novel End-to-end Framework for Conversational Machine Reading Comprehension (2022.coling-1)

Challenge:	composition of pre-training datasets for large language models remains undisclosed . current methods for evaluating data quality are limited by single-dimensional evaluation or redundancy-focused strategies.
Approach:	They propose a multi-dimensional data selection method that integrates dimensions with existing quality metrics through learned optimal weightings.
Outcome:	The proposed method doubles convergence speed for 1.3B model models and improves downstream task performance by 3.23%.

Copied to clipboard

Xiao Zhang, Heyan Huang, Zewen Chi, Xian-Ling Mao

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

Challenge:	Existing methods require three steps to understand text, but span extraction and question rephrasing steps are not fully exploited.
Approach:	They propose a framework for conversational machine reading comprehension based on shared parameter mechanism . experimental results show the proposed framework achieves new state-of-the-art results on the ShARC leaderboard .
Outcome:	The proposed framework achieves state-of-the-art on the ShARC leaderboard with the BLEU-4 score of 55.2.

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

RSMeM: Knowledge-Enhanced Memory Evolution for Remote Sensing Agents with Systematic Evaluation (2026.acl-long)

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

Bingxian Wu, Yu Zhang, Zonghao Guo, Tang Liu, Chen Qian, Yuxiang Lu, Xingbo Du, Yanghao Li, Yidan Zhang, Chi Chen, Ling Yao, Chenghu Zhou, Maosong Sun

Cross-Lingual Phrase Retrieval (2022.acl-long)

Challenge:	Existing RS agents built on general-purpose LLMs are domain-agnostic, resulting in brittle and error-prone workflows.
Approach:	They propose a knowledge-enhanced memory evolution mechanism that bootstraps RS agents with pre-distilled domain knowledge and iteratively integrates online experience for robust multi-step tool execution.
Outcome:	Experiments show that the new model improves tool-use performance and accuracy . iteratively, iteration of the model integrates online experience for robust multi-step tool execution .

Copied to clipboard

Heqi Zheng, Xiao Zhang, Zewen Chi, Heyan Huang, Yan Tan, Tian Lan, Wei Wei, Xian-Ling Mao

Cross-Lingual Transfer

Challenge:	Existing approaches to cross-lingual phrase retrieval learn word or sentence representations in word or sentences.
Approach:	They propose a cross-lingual phrase retrieval model that extracts phrase representations from unlabeled example sentences.
Outcome:	The proposed model outperforms state-of-the-art methods on a large-scale cross-lingual phrase retrieval dataset, showing it can perform in an unseen language pair during training.

ActiView: Evaluating Active Perception Ability for Multimodal Large Language Models (2025.acl-long)

Copied to clipboard

Ziyue Wang, Chi Chen, Fuwen Luo, Yurui Dong, Yuanchi Zhang, Yuzhuang Xu, Xiaolong Wang, Peng Li, Yang Liu

Leveraging LLM Reasoning Enhances Personalized Recommender Systems (2024.findings-acl)

Challenge:	Existing benchmarks for evaluating MLLMs have not addressed active perception . a novel benchmark is proposed to evaluate active perception in ML models .
Approach:	They propose a benchmark to evaluate active perception in Multimodal Large Language Models . they restrict the perceptual field of a model and require it to actively zoom or shift it .
Outcome:	The proposed benchmark focuses on a specialized form of Visual Question Answering (VQA) that eases and quantifies the evaluation yet challenging for existing MLLMs.

Copied to clipboard

Alicia Tsai, Adam Kraft, Long Jin, Chenwei Cai, Anahita Hosseini, Taibai Xu, Zemin Zhang, Lichan Hong, Ed H. Chi, Xinyang Yi

LaMP-Val: Large Language Models Empower Personalized Valuation in Auction (2025.findings-emnlp)

Challenge:	Recent advances have showcased the potential of Large Language Models (LLMs) in executing reasoning tasks, particularly facilitated by Chain-of-Thought (CoT) prompting.
Approach:	They propose to use Large Language Models to perform tasks with subjectivity and personalized preferences as inputs to RecSys.
Outcome:	The proposed framework aligns with real human judgment on the coherence and faithfulness of LLM reasoning responses.

Copied to clipboard

Jie Sun, Tianyu Zhang, Houcheng Jiang, Kexin Huang, Xiang Shu, Zhibo Zhu, Lintao Ma, Xingyu Lu, Jun Zhou, Junkang Wu, Chi Luo, An Zhang, Jiancan Wu, Xiang Wang

Demystifying Multi-Agent Debate: The Role of Confidence and Diversity (2026.findings-acl)

Challenge:	Currently, most research focuses on the bidding algorithms used within auction mechanisms.
Approach:	They propose a personalized valuation framework that integrates Large Language Models to incorporate personalized semantic preference into users valuation process.
Outcome:	The proposed framework incorporates Large Language Models to incorporate personalized semantic preference into users valuation process.

Copied to clipboard

Xiaochen Zhu, Caiqi Zhang, Yizhou Chi, Tom Stafford, Nigel Collier, Andreas Vlachos

Challenge:	Multi-agent debate (MAD) is widely used to improve large language models' (LLMs) reasoning and test-time scaling.
Approach:	They propose a diversity-aware initialisation that selects a more diverse pool of candidate answers, increasing the likelihood that a correct hypothesis is present at the start of debate.
Outcome:	The proposed protocol outperforms vanilla MAD and majority vote on six reasoning-oriented QA benchmarks.

DLTKG: Denoising Logic-based Temporal Knowledge Graph Reasoning (2025.findings-emnlp)

Copied to clipboard

Xiaoke Wang, Fu Zhang, Jingwei Cheng, Yiwen Chi, Jiashun Peng, Yingsong Ning

Knowledge Representation

Challenge:	Current approaches to temporal knowledge representation face limited generalization to unseen facts and insufficient interpretability of reasoning processes.
Approach:	They propose a framework that uses a denoising diffusion process to complete reasoning tasks . they propose introducing a noise source and historical conditionguiding mechanism to improve interpretability .
Outcome:	The proposed framework outperforms state-of-the-art methods on three benchmark datasets.

ProtLLM: An Interleaved Protein-Language LLM with Protein-as-Word Pre-Training (2024.acl-long)

Copied to clipboard

Le Zhuo, Zewen Chi, Minghao Xu, Heyan Huang, Jianan Zhao, Heqi Zheng, Conghui He, Xian-Ling Mao, Wentao Zhang

Browse and Concentrate: Comprehending Multimodal Content via Prior-LLM Context Fusion (2024.acl-long)

Challenge:	Experimental results demonstrate that ProtLLM achieves superior performance against protein-specialized baselines on protein-centric tasks and induces zero-shot and in-context learning capabilities on protein language tasks.
Approach:	They propose a cross-modal large language model (LLM) that can handle protein-centric and protein-language tasks by using a dynamic protein mounting mechanism.
Outcome:	The proposed model can predict proteins from a vast pool of candidates and can also predict natural language and biological papers.

Copied to clipboard

Ziyue Wang, Chi Chen, Yiqi Zhu, Fuwen Luo, Peng Li, Ming Yan, Ji Zhang, Fei Huang, Maosong Sun, Yang Liu

Why Does New Knowledge Create Messy Ripple Effects in LLMs? (2024.emnlp-main)

Challenge:	Multimodal Large Language Models (MLLMs) lack understanding of multi-image and interleaved inputs due to the visual features encoded by frozen encoders before being fed into the LLM backbone.
Approach:	They propose a two phase paradigm to enable in-depth multimodal context fusion prior to feeding the features into LLMs.
Outcome:	The proposed paradigm boosts the performance on 7 multi-image scenarios, contributing to increments on average accuracy by 2.13% and 7.60% against strong MLLMs baselines with 3B and 11B LLMs, respectively.

Copied to clipboard

Jiaxin Qin, Zixuan Zhang, Chi Han, Pengfei Yu, Manling Li, Heng Ji

AgentCPM-GUI: Building Mobile-Use Agents with Reinforcement Fine-Tuning (2025.emnlp-demos)

Challenge:	Existing research has focused on post-training knowledge editing (KE) for language models to ensure that knowledge remains accurate and up-to-date.
Approach:	They propose to use a GradSim indicator to detect when and why updated knowledge ripples in language models.
Outcome:	The proposed indicator GradSim shows that LMs that fail to handle ripple effects have low GradSIM.

Copied to clipboard

Zhong Zhang, Yaxi Lu, Yikun Fu, Yupeng Huo, Shenzhi Yang, Yesai Wu, Han Si, Xin Cong, Haotian Chen, Yankai Lin, Xie Xie, Wei Zhou, Wang Xu, Zhou Su, Zhongwu Zhai, Xiaoming Liu, null Meiyudong, Jianming Xu, Hongyan Tian, Chongyi Wang, Chi Chen, Yuan Yao, Zhiyuan Liu, Maosong Sun

Bridging The Gap: Entailment Fused-T5 for Open-retrieval Conversational Machine Reading Comprehension (2023.acl-long)

Challenge:	Large language model agents have enabled GUI-based automation, but their deployment is limited by noisy data, poor generalization, and lack of support for non-English GUIs.
Approach:	They propose an 8B-parameter GUI agent built for robust and efficient on-device GUI interaction.
Outcome:	The proposed GUI agent achieves promising performance on five public benchmarks and proposed Chinese benchmark CAGUI.

Copied to clipboard

Xiao Zhang, Heyan Huang, Zewen Chi, Xian-Ling Mao

Natural Language Interfaces

Challenge:	Open-retrieval conversational machine reading comprehension (OCMRC) simulates real-life conversation scenes.
Approach:	They propose a one-stage end-to-end framework to bridge the information gap between decision-making and question generation in a global understanding manner.
Outcome:	The proposed framework achieves new state-of-the-art performance on the OR-ShARC benchmark.

Bridging the Granularity Gap for Acoustic Modeling (2023.findings-acl)

Copied to clipboard

Chen Xu, Yuhao Zhang, Chengbo Jiao, Xiaoqian Liu, Chi Hu, Xin Zeng, Tong Xiao, Anxiang Ma, Huizhen Wang, Jingbo Zhu

Efficient Pretraining Data Selection for Language Models via Multi-Actor Collaboration (2025.acl-long)

Challenge:	Despite the success of speech recognition, how to encode the speech features effectively remains an open problem.
Approach:	They propose a Progressive Down-Sampling technique which compresses acoustic features into coarser-grained units containing more complete semantic information, like text-level representation.
Outcome:	The proposed method yields comparable or better results on the speech recognition task and inference speedups ranging from 1.20x to 1.47x.

Copied to clipboard

Tianyi Bai, Ling Yang, Zhen Hao Wong, Fupeng Sun, Xinlin Zhuang, Jiahui Peng, Chi Zhang, Lijun Wu, Qiu Jiantao, Wentao Zhang, Binhang Yuan, Conghui He

Code over Words: Overcoming Semantic Inertia via Code-Grounded Reasoning (2026.findings-acl)

Challenge:	Efficient data selection is crucial to accelerate the pretraining of language models . limited research has addressed the inherent conflicts between data selection methods .
Approach:	They propose a multi-actor collaborative data selection mechanism that prioritizes data based on its specific criterion and updates prioritization rules using the current state of the model.
Outcome:	The proposed model accelerates convergence in LM pretraining and achieves an average relative performance gain of 10.5% across multiple language model benchmarks.

Copied to clipboard

Manjie Xu, Isabella Yin, Xinyi Tu, Chi Zhang, Yixin Zhu

Do Images Speak Louder than Words? Investigating the Effect of Textual Misinformation in VLMs (2026.eacl-long)

Challenge:	Large Language Models struggle with semantic inertia, a problem that is often attributed to natural language encoding, which entangles descriptive semantics and logical rules, leading to persistent hallucinations of familiar physics despite explicit contradictory rules.
Approach:	They propose a framework that decouples logical dynamics from visual priors via amortized theory induction and counterfactual contrastive alignment.
Outcome:	The proposed framework outperforms expensive inference-time search methods in both efficiency and accuracy.

Copied to clipboard

Chi Zhang, Wenxuan Ding, Jiale Liu, Mingrui Wu, Qingyun Wu, Ray Mooney

Responsible & Trustworthy NLP

Challenge:	Existing studies have shown that Vision-Language Models have robust multimodal reasoning capabilities, but their robustness against textual misinformation remains under-explored.
Approach:	They propose to use visual-question-answering (VQA) prompts to generate persuasive prompts that deliberately conflict with visual evidence to test their models.
Outcome:	The proposed framework shows that models are vulnerable to misleading prompts, and show an average performance drop of over 48.2% after only one round of persuasive conversation.

A Multi-Task Learning Framework for Extracting Bacteria Biotope Information (D19-57)

Copied to clipboard

Qi Zhang, Chao Liu, Ying Chi, Xuansong Xie, Xiansheng Hua

PESCO: Prompt-enhanced Self Contrastive Learning for Zero-shot Text Classification (2023.acl-long)

Challenge:	Existing methods to extract information from unstructured text are slow or expensive to get.
Approach:	They propose a multi-task transfer multi-learning method for Bacteria Biotope rel+ner task . they use BERT and pre-train it using mask language models and next sentence prediction .
Outcome:	The proposed method achieves the best performance on all metrics including slot error rate, precision and recall in the Bacteria Biotope rel+ner subtask.

Copied to clipboard

Yau-Shian Wang, Ta-Chung Chi, Ruohong Zhang, Yiming Yang

Clustering and Ranking: Diversity-preserved Instruction Selection through Expert-aligned Quality Estimation (2024.emnlp-main)

Challenge:	Existing text classification frameworks require large amounts of human-labeled documents to train .
Approach:	They propose a contrastive learning framework that improves zero-shot text classification . they add prompts to enhance label retrieval and use retrieved labels to enrich training .
Outcome:	The proposed framework achieves state-of-the-art on four benchmark text classification datasets.

Copied to clipboard

Yuan Ge, Yilun Liu, Chi Hu, Weibin Meng, Shimin Tao, Xiaofeng Zhao, Mahong Xia, Zhang Li, Boxing Chen, Hao Yang, Bei Li, Tong Xiao, JingBo Zhu

InferPilot: Autonomous Inference Attacks Against ML Services With LLM-Based Agents (2026.findings-acl)

Challenge:	Existing methods for instruction data selection have limitations such as relying on fragile external APIs, being affected by biases in GPT models, or reducing the diversity of the selected instruction dataset.
Approach:	They propose an industrial-friendly, expert-aligned and diversity-preserved instruction data selection method: Clustering and Ranking (CaR).
Outcome:	The proposed method outperforms Alpaca's existing methods by 32.1% in GPT-4 evaluations.

Copied to clipboard

Yixin Wu, Rui Wen, Chi Cui, Michael Backes, Yang Zhang

Robustness in NLP

Challenge:	Inference attacks are important for assessing model's robustness, but their implementation and parameters are challenging for non-experts.
Approach:	They propose an autonomous agent capable of conducting inference attacks without human intervention.
Outcome:	The proposed agent achieves a 100.0% task completion rate and near-expert attack performance with an average token cost of only 0.627 per run.

SWE-QA-Pro: A Representative Benchmark and Scalable Training Recipe for Repository-Level Code Understanding (2026.findings-acl)

Copied to clipboard

Songcheng Cai, Zhiheng Lyu, Yuansheng Ni, Xiangchao Chen, Baichuan Zhou, Shenzhe Zhu, Yi Lu, Haozhe Wang, Chi Ruan, Benjamin Schneider, Weixu Zhang, Xiang Li, Andy Zheng, Yuyu Zhang, Ping Nie, Wenhu Chen

Spectral Characterization and Mitigation of Sequential Knowledge Editing Collapse (2026.acl-long)

Challenge:	Existing benchmarks for agentic repository-level code understanding overlook long tail topics and rely on memorized knowledge.
Approach:	They propose a repository-level agentic code understanding benchmark that uses long-tail repositories with executable environments to enforce topical balance.
Outcome:	Empirically, a Qwen3-8B model trained with the proposed benchmark outperforms GPT-4o by 2.3 points.

Copied to clipboard

Chi Zhang, Mengqi Zhang, Xiaotian Ye, Runxi Cheng, Zisheng Zhou, Ying Zhou, Pengjie Ren, Zhumin Chen

Challenge:	Existing approaches to reducing the effects of knowledge editing are insufficiently understood.
Approach:	They propose a plug-and-play framework that preserves the dominant subspace of the original weights and analyzes parameter updates in the spectral basis of the weights.
Outcome:	The proposed framework improves editing efficacy while preserving general abilities under long-horizon sequential editing, including extreme settings with up to 20,000 edits.

Omni-I2C: A Holistic Benchmark for High-Fidelity Image-to-Code Generation (2026.acl-long)

Copied to clipboard

Jiawei Zhou, Chi Zhang, Xiang Feng, Qiming Zhang, Haibo Qiu, Lihuo He, Dengpan Ye, Xinbo Gao, Jing Zhang

Dynamic Infilling Anchors for Format-Constrained Generation in Diffusion Large Language Models (2026.acl-long)

Challenge:	a benchmark is designed to evaluate the capability of Large Multimodal Models (LMMs) in converting complex, structured digital graphics into executable code.
Approach:	They propose a benchmark to evaluate the capability of Large Multimodal Models to convert digital graphics into executable code.
Outcome:	The proposed benchmark exposes the performance gap among leading LMMs . the benchmark features 1130 meticulously curated samples .

Copied to clipboard

Boyan Han, Yiwei Wang, Yi Song, Yujun Cai, Chi Zhang

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

Challenge:	Diffusion large language models (dLLMs) offer bidirectional attention and parallel generation . fixed anchors can enforce constraints, but they often impose rigid spans, leading to truncated reasoning .
Approach:	They propose a method that dynamically estimates end-anchor positions to adjust generation length before iterative infilling.
Outcome:	The proposed method improves format compliance and answer accuracy on GSM8K and MATH.

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

Benchmarking Large Vision-Language Models on CFMME: A Comprehensive Chinese Financial Multimodal Evaluation Dataset (2026.acl-long)

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

Qian Chen, Xianyin Zhang, Yanzhi Liu, Lifan Guo, Feng Chen, Chi Zhang

RENN: A Rule Embedding Enhanced Neural Network Framework for Temporal Knowledge Graph Completion (2024.lrec-main)

Challenge:	Large Vision-Language Models (LVLMs) have expanded capabilities beyond text understanding . a novel Chinese financial multimodal evaluation benchmark is used to evaluate LVLM capabilities .
Approach:	They propose a Chinese financial multimodal evaluation benchmark to evaluate LVLMs' capabilities . the model has an overall accuracy of 66.11% and an average score of 77.18 .
Outcome:	The proposed model achieves an overall accuracy of 66.11% on the question answering task and an average score of 77.18 on detection, recognition, and information extraction tasks.

Copied to clipboard

Linlin Zong, Zhenrong Xie, Chi Ma, Xinyue Liu, Xianchao Zhang, Bo Xu

FastSeq: Make Sequence Generation Faster (2021.acl-demo)

Challenge:	Existing methods for temporal knowledge graph embedding do not account for structural dependencies between relations.
Approach:	They propose a framework that enhances temporal knowledge graph completion through rule embedding.
Outcome:	The proposed framework improves temporal knowledge graph completion through rule embedding.

Copied to clipboard

Yu Yan, Fei Hu, Jiusheng Chen, Nikhil Bhendawade, Ting Ye, Yeyun Gong, Nan Duan, Desheng Cui, Bingyu Chi, Ruofei Zhang

Synchronous Double-channel Recurrent Network for Aspect-Opinion Pair Extraction (2020.acl-main)

Challenge:	Transformer-based models have made tremendous impact in natural language generation, but inference speed is still a bottleneck due to large model size and intensive computing involved in auto-regressive decoding process.
Approach:	They propose an attention cache optimization, an efficient algorithm for detecting repeated n-grams, and an asynchronous generation pipeline with parallel I/O to accelerate sequence generation without loss of accuracy.
Outcome:	The proposed framework can accelerate the sequence generation by 4x to 9x with a simple one-line code change for a set of widely used and diverse models.

Copied to clipboard

Shaowei Chen, Jie Liu, Yu Wang, Wenzheng Zhang, Ziming Chi

Information Extraction & Text Mining

Challenge:	Existing studies focus on aspect-opinion relation detection, but neglect to recognize the relations between aspects and opinion expressions.
Approach:	They propose a Synchronous Double-channel Recurrent Network to deal with AOPE task . they propose an opinion entity extraction unit, a relation detection unit, and a synchronization unit .
Outcome:	The proposed system achieves state-of-the-art in opinion entity extraction . it is based on three datasets based upon SemEval 2014 and 2015 benchmarks .

MUSEG: Reinforcing Video Temporal Understanding via Timestamp-Aware Multi-Segment Grounding (2026.acl-long)

Copied to clipboard

Fuwen Luo, Shengfeng Lou, Chi Chen, Ziyue Wang, Chenliang Li, Weizhou Shen, Jiyue Guo, Peng Li, Ming Yan, Ji Zhang, Fei Huang, Yang Liu

Visual Data in NLP

Challenge:	Existing methods for MLLMs struggle with fine-grained temporal reasoning . despite advances in video understanding, current methods struggle with time-sensitive tasks .
Approach:	They propose a time-stamp-aware multi-segment grounding method that enhances temporal understanding by introducing timestamps.
Outcome:	The proposed method outperforms existing methods on time-sensitive tasks and generalizes well across diverse temporal understanding scenarios.

STRICT: Stress-Test of Rendering Image Containing Text (2025.emnlp-main)

Copied to clipboard

Tianyu Zhang, Xinyu Wang, Lu Li, Zhenghan Tai, Jijun Chi, Jingrui Tian, Hailin He, Suyuchen Wang

Learning Architectures from an Extended Search Space for Language Modeling (2020.acl-main)

Challenge:	Despite the advances in diffusion models, the generation of coherent text remains a major bottleneck.
Approach:	They propose a benchmark to test the ability of diffusion models to render coherent text in images.
Outcome:	The proposed model fails to generate coherent and legible text in images despite its iterative nature . the model fails in both the maximum length of readable text and correctness and legibility of the generated text .

Copied to clipboard

Yinqiao Li, Chi Hu, Yuhao Zhang, Nuo Xu, Yufan Jiang, Tong Xiao, Jingbo Zhu, Tongran Liu, Changliang Li

Enhanced Visual Instruction Tuning with Synthesized Image-Dialogue Data (2024.findings-acl)

Challenge:	Neural architecture search (NAS) has advanced in recent years but most NAS systems restrict search to learning architectures of a recurrent or convolutional cell.
Approach:	They propose a general approach to learn both intra-cell and inter-cell architectures . they implement their approach in a differentiable architecture search system .
Outcome:	The proposed approach outperforms the baseline on PTB and WikiText data and shows good transferability to other systems.

Copied to clipboard

Yanda Li, Chi Zhang, Gang Yu, Wanqi Yang, Zhibin Wang, Bin Fu, Guosheng Lin, Chunhua Shen, Ling Chen, Yunchao Wei

Towards Stable and Effective Reinforcement Learning for Mixture-of-Experts (2026.acl-long)

Challenge:	OpenAI's GPT-4 has demonstrated remarkable multimodal capabilities, but specific mechanics of GPT4 remain unknown.
Approach:	They propose a data collection methodology that synchronously synthesizes images and dialogues for visual instruction tuning.
Outcome:	The proposed method improves on ten commonly assessed models and provides greater flexibility compared to existing methods.

Copied to clipboard

Di Zhang, Xun Wu, Shaohan Huang, Lingjie Jiang, Yaru Hao, Li Dong, Zewen Chi, Zhifang Sui, Furu Wei

Divide, Optimize, Merge: Scalable Fine-Grained Generative Optimization for LLM Agents (2025.findings-emnlp)

Challenge:	Reinforcement learning with verifiable rewards (RLVR) training with Mixture-of-Experts policies remains fragile and prone to reward collapse.
Approach:	They propose a router shift-based policy optimization method that computes a per-token router-shift ratio conditioned on the previously activated experts and applies stop-gradient and a lower-bound floor.
Outcome:	The proposed method achieves better performance and greater stability than previous methods.

Copied to clipboard

Jiale Liu, Yifan Zeng, Shaokun Zhang, Chi Zhang, Malte Højmark-Bertelsen, Marie Normann Gadeberg, Huazheng Wang, Qingyun Wu

Rethinking Assessments of Prompt Injection Attacks (2026.findings-acl)

Challenge:	LLM-based generative optimization has shown remarkable potential in improving agentic systems, but the current approach of prompting with the trajectories on the whole training dataset becomes untenable as datasets grow.
Approach:	They propose a scalable framework that divides large optimization tasks into manageable subsets and performs targeted optimizations.
Outcome:	The proposed framework outperforms conventional approach by 1.6-8.6% while reducing average prompt token consumption by 56.3%.

Copied to clipboard

Chi Cui, Yixin Wu, Michael Backes, Yang Zhang

Responsible & Trustworthy NLP

Challenge:	Prompt injection attacks are recognized as one of the primary risks faced by LLM-integrated applications in recent years.
Approach:	They evaluate prompt injection attacks on LLM-integrated applications across 37 target tasks, 185 injected tasks, 21 attack instructions, and 143,745 queries.
Outcome:	The proposed framework provides a solid foundation for assessing vulnerabilities in LLM-integrated applications and evaluating the efficacy of defensive strategies.

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

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.

SQLAgent: Learning to Explore Before Generating as a Data Engineer (2026.findings-acl)

Copied to clipboard

Wenjia Jiang, Yiwei Wang, Boyan Han, Joey Tianyi Zhou, Chi Zhang

Programming Languages in NLP

Challenge:	Existing approaches to large language models fail to generalize in complex, real-world settings due to database-specific nature of SQL reasoning.
Approach:	They propose a two-stage LLM-based framework that decouples knowledge acquisition from query generation.
Outcome:	The proposed framework significantly improves accuracy over baselines on large-scale benchmarks.

Hard to Read, Easy to Jailbreak: How Visual Degradation Bypasses MLLM Safety Alignment (2026.findings-acl)

Copied to clipboard

Zhixue Song, Boyan Han, Yiwei Wang, Chi Zhang

The Law of Knowledge Overshadowing: Towards Understanding, Predicting and Preventing LLM Hallucination (2025.findings-acl)

Challenge:	Recent advances in visual context compression enable MLLMs to process ultra-long contexts efficiently by rendering text into images.
Approach:	They propose a strategy that decouples visual transcription from safety auditing by enforcing a serialized pipeline to decouple visual transcription and safety assessment.
Outcome:	The proposed strategy decouples visual transcription from safety auditing to reduce the risk of jailbreaking.

Copied to clipboard

Yuji Zhang, Sha Li, Cheng Qian, Jiateng Liu, Pengfei Yu, Chi Han, Yi R. Fung, Kathleen McKeown, ChengXiang Zhai, Manling Li, Heng Ji

Improved Differentiable Architecture Search for Language Modeling and Named Entity Recognition (D19-1)

Challenge:	Hallucination is a persistent challenge in large language models where even with rigorous quality control, models often generate distorted facts.
Approach:	They propose a new framework to quantify factual hallucinations by modeling knowledge overshadowing.
Outcome:	The proposed framework improves model factuality on Overshadow (27.9%), MemoTrap (13.1%) and NQ-Swap (18.3%).

Copied to clipboard

Yufan Jiang, Chi Hu, Tong Xiao, Chunliang Zhang, Jingbo Zhu

Hiring Now: A Skill-Aware Multi-Attention Model for Job Posting Generation (2020.acl-main)

Challenge:	Neural architecture search (NAS) is a popular approach for finding new models and freeing researchers from the hard work of designing network architectures.
Approach:	They propose differentiable neural architecture search methods for natural language processing . they remove the softmax-local constraint and apply it to named entity recognition .
Outcome:	The proposed method outperforms strong baselines on the language modeling task.

Copied to clipboard

Liting Liu, Jie Liu, Wenzheng Zhang, Ziming Chi, Wenxuan Shi, Yalou Huang

Challenge:	Creating job requirements is a crucial step in the recruiting process, but it is difficult to specify the level of education, experience, relevant skills per the job description.
Approach:	They propose a conditional text generation task to generate job requirements based on job descriptions . they use a hierarchical decoder to label the job description with multiple skills . a skill knowledge graph is constructed to capture the global prior knowledge about skills based upon the model .
Outcome:	The proposed method is evaluated on real-world job posting data.

An Automatic Method to Estimate Correctness of RAG (2025.coling-industry)

Copied to clipboard

Chi Zhang, Vivek V. Datla, Aditya Shrivastava, Alfy Samuel, Zhiqi Huang, Anoop Kumar, Daben Liu

Challenge:	Existing methods to assess the correctness of RAG models fail to capture the model’s internal state during answer generation.
Approach:	They propose a method to predict the correctness of RAG models by modeling the model’s uncertainty on quantified perturbations of input.
Outcome:	Extensive experiments across multiple large language models show that the proposed approach quantifies RAG robustness by aligning predictions with ground truth with a MSE 0.002 while offering flexibility for diverse qualitative metrics.

Prompt Combines Paraphrase: Teaching Pre-trained Models to Understand Rare Biomedical Words (2022.coling-1)

Copied to clipboard

Haochun Wang, Chi Liu, Nuwa Xi, Sendong Zhao, Meizhi Ju, Shiwei Zhang, Ziheng Zhang, Yefeng Zheng, Bing Qin, Ting Liu

FinHEAR: Human Expertise and Adaptive Risk-Aware Temporal Reasoning for Financial Decision-Making (2025.findings-emnlp)

Challenge:	Pre-trained models perform poorly with limited data and rare biomedical words.
Approach:	They propose to use prompt to fine-tune pre-trained models for biomedical domain tuning with a simple approach.
Outcome:	The proposed method achieves up to 6% improvement in biomedical natural language inference task without any extra parameters or training steps using few-shot vanilla prompt settings.

Copied to clipboard

Jiaxiang Chen, Mingxi Zou, Zhuo Wang, Qifan Wang, Danny Dongning Sun, Zhang Chi, Zenglin Xu