Papers by Yang Chi

RankPrompt: Step-by-Step Comparisons Make Language Models Better Reasoners (2024.lrec-main)

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

Chi Hu, Yuan Ge, Xiangnan Ma, Hang Cao, Qiang Li, Yonghua Yang, Tong Xiao, Jingbo Zhu

Dialogue Systems & Conversational Agents

Challenge:	Existing solutions to reasoning tasks require extensive human annotations or fail in scenarios with inconsistent responses.
Approach:	They propose a new method that enables LLMs to self-rank their responses without additional resources.
Outcome:	The proposed method improves reasoning performance of ChatGPT and GPT-4 with 13% improvement over existing methods.

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

Copied to clipboard

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

ETR: Entropy Trend Reward for Efficient Chain-of-Thought Reasoning (2026.acl-long)

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

Xuan Xiong, Huan Liu, Li Gu, Zhixiang Chi, Yue Qiu, Yuanhao YU, Yang Wang

Responsible & Trustworthy NLP

Challenge:	Existing methods to shorten CoTs use length penalties or global entropy reduction . Existing approaches to CoT reasoning have significant practical drawbacks .
Approach:	They propose a method that shortens CoTs by length penalties or global entropy reduction . they integrate ETR into Group Relative Policy Optimization and evaluate it .
Outcome:	The proposed objective improves accuracy–efficiency trade-off by +9.9% while reducing CoT length by 67% across four benchmarks.

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

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

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

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

Can Language Models Follow Multiple Turns of Entangled Instructions? (2025.findings-emnlp)

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

Chi Han, Xin Liu, Haodong Wang, Shiyang Li, Jingfeng Yang, Haoming Jiang, Zhengyang Wang, Qingyu Yin, Liang Qiu, Changlong Yu, Yifan Gao, Zheng Li, Bing Yin, Jingbo Shang, Heng Ji

ThoughtSculpt: Reasoning with Intermediate Revision and Search (2025.findings-naacl)

Challenge:	Despite of significant achievements in improving instruction-following capabilities of large language models, the ability to process multiple potentially entangled or conflicting instructions remains a considerable challenge.
Approach:	They construct multi-turn instruction with 1.1K high-quality multi-turned conversations using the human-in-the-loop approach and examine their capabilities.
Outcome:	The proposed model shows that it is difficult to integrate multiple turns and balance competing objectives when instructions intersect or conflict.

Copied to clipboard

Yizhou Chi, Kevin Yang, Dan Klein

Reasoning

Challenge:	THOUGHTSCULPT is a general reasoning and search method for tasks with outputs that can be decomposed into components.
Approach:	They propose a general reasoning and search method for tasks with outputs that can be decomposed into components.
Outcome:	THOUGHTSCULPT outperforms state-of-the-art reasoning methods on three tasks . authors show that distinct prompting strategies can influence the performance of LLMs .

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

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

End-to-End Unsupervised Vision-and-Language Pre-training with Referring Expression Matching (2022.emnlp-main)

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

Chi Chen, Peng Li, Maosong Sun, Yang Liu

Ancient Chinese Glyph Identification Powered by Radical Semantics (2024.findings-acl)

Challenge:	Existing unsupervised vision-and-language pre-training methods take pre-extracted region-based visual features from external object detectors, which limits flexibility and reduces computational efficiency.
Approach:	They propose an unsupervised vision-and-language pre-training task that predicts which patches contain an object referred to in natural language from the encoded visual features.
Outcome:	The proposed approach outperforms existing methods and obtains state-of-the-art results on four vision-and-language tasks.

Copied to clipboard

Yang Chi, Fausto Giunchiglia, Chuntao Li, Hao Xu

Mask-Align: Self-Supervised Neural Word Alignment (2021.acl-long)

Challenge:	Currently, about half of ancient Chinese glyphs have not been deciphered yet.
Approach:	They propose to use a Chinese glyph knowledge graph to infer the Chinese character label for the unknown ancient Chinese . they propose to combine the visual, textual, and the graph data to create a multimodal Chinese morph identification framework.
Outcome:	The proposed method can identify ancient Chinese characters from 1300 BC to 200 BC based on image and radical semantics on a 1000-year-old Chinese glyph dataset.

Copied to clipboard

Chi Chen, Maosong Sun, Yang Liu

Machine Translation

Challenge:	Word alignment is an important task in many natural language processing tasks.
Approach:	They propose a self-supervised word alignment model that takes advantage of the full context on the target side.
Outcome:	The proposed model outperforms previous unsupervised models and obtains state-of-the-art results on four language pairs.

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

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.

Sparse Brains are Also Adaptive Brains: Cognitive-Load-Aware Dynamic Activation for LLMs (2026.findings-eacl)

Copied to clipboard

Yiheng Yang, Yujie Wang, Chi Ma, Lei Yu, Emmanuele Chersoni, Chu-Ren Huang

Green & Sustainable NLP

Challenge:	Existing sparsity methods lack adaptivity to contextual or model structural demands or incur prohibitive computational overhead.
Approach:	They propose a Cognitive-Load-Aware Dynamic Activation framework that synergizes statistical sparsity with semantic adaptability.
Outcome:	The proposed framework achieves 20% average speedup with less than 2% accuracy degradation outperforming Griffin and TT.

Knowledge-Infused Multi-Bit Watermarking for RAG Knowledge Bases (2026.findings-acl)

Copied to clipboard

Ke Yang, Shuguang Yuan, Jing Yu, Chi Chen

ChartEdit: How Far Are MLLMs From Automating Chart Analysis? Evaluating MLLMs’ Capability via Chart Editing (2025.findings-acl)

Challenge:	Existing RAG watermarking methods are limited in their encoding capacity and potential degradation of performance or knowledge quality.
Approach:	They propose knowledge-infused and multi-bit watermarking (KMW) for RAG knowledge bases by benign knowledge completion and a tailored generative watermark algorithm.
Outcome:	The proposed method extracts watermarks from adversarial RAGs while remaining stealthy and secure.

Copied to clipboard

Xuanle Zhao, Xuexin Liu, Yang Haoyue, Xianzhen Luo, Fanhu Zeng, Jianling Li, Qi Shi, Chi Chen

CODIS: Benchmarking Context-dependent Visual Comprehension for Multimodal Large Language Models (2024.acl-long)

Challenge:	Existing evaluations of multimodal large language models rely on limited case studies . however, they lack the ability to generate accurate edits according to the instructions .
Approach:	They propose a benchmark for chart editing that includes 1,405 edit instructions applied to 233 real-world charts.
Outcome:	The proposed benchmark includes 1,405 diverse editing instructions applied to 233 real-world charts.

Copied to clipboard

Fuwen Luo, Chi Chen, Zihao Wan, Zhaolu Kang, Qidong Yan, Yingjie Li, Xiaolong Wang, Siyu Wang, Ziyue Wang, Xiaoyue Mi, Peng Li, Ning Ma, Maosong Sun, Yang Liu

You Can Have a Second Chance: Unbiased and Multi-bit Watermarking for Diffusion Language Models with Regret-based Remasking (2026.acl-long)

Challenge:	Multimodal large language models have demonstrated promising results in a variety of tasks that combine vision and language.
Approach:	They propose a benchmark to assess the ability of models to use contextual information in free-form text to enhance visual comprehension.
Outcome:	The proposed model fails to extract and utilize contextual information to improve understanding of images.

Copied to clipboard

Ke Yang, Dongyang Liang, Jing Yu, Shuguang Yuan, Chi Chen

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

Challenge:	Existing sequential LLMs cannot be directly applied to DLMs, as their generation order is arbitrary.
Approach:	They propose a stability-aware constraint that allows watermarking only in stable contexts and a bit-controlled, unbiased modulation to preserve the original DLM output distribution.
Outcome:	The proposed scheme achieves stable watermarking with minimal quality impact while maintaining high detection accuracy and multi-bit capacity.

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.

ZiNet: Linking Chinese Characters Spanning Three Thousand Years (2022.findings-acl)

Copied to clipboard

Yang Chi, Fausto Giunchiglia, Daqian Shi, Xiaolei Diao, Chuntao Li, Hao Xu

Knowledge Representation

Challenge:	tens of thousands of ancient characters must be deciphered by experts to interpret unearthed documents.
Approach:	They propose a diachronic Chinese knowledge base to help researchers discover glyph similar characters by measuring glyph similarities between ancient Chinese characters.
Outcome:	The proposed method shows strong correlations between the scores obtained from the method and from human experts.

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

Copied to clipboard

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

Beyond Static Rules: Automated Discovery of Latent Vulnerabilities in Text-to-SQL (2026.findings-acl)

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

Hanqing Wang, Yongdong Chi, Jian Yang, Lei Yang, Jiehui Zhao, Yun Chen, Guanhua Chen

Programming Languages in NLP

Challenge:	Large Language Models (LLMs) have been successful in Text-to-SQL tasks, but their deployment in real-world environments is hindered by latent reliability issues.
Approach:	They propose a framework to autonomously uncover latent failure patterns in LLM-based Text-to-SQL generation.
Outcome:	The proposed framework uncovers a substantial number of failure cases on state-of-the-art open-source LLMs.

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

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.

Filling the Image Information Gap for VQA: Prompting Large Language Models to Proactively Ask Questions (2023.findings-emnlp)

Copied to clipboard

Ziyue Wang, Chi Chen, Peng Li, Yang Liu

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

Challenge:	Large Language Models (LLMs) demonstrate impressive reasoning ability and the maintenance of world knowledge in natural language tasks.
Approach:	They propose a framework that enables LLMs to ask relevant questions to uncover more details in the image, along with filters for refining the generated information.
Outcome:	The proposed framework boosts the performance of baseline methods by 2.15% on OK-VQA and achieves consistent improvements across different LLMs.

Copied to clipboard

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

Weakly Supervised Vision-and-Language Pre-training with Relative Representations (2023.acl-long)

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

Copied to clipboard

Chi Chen, Peng Li, Maosong Sun, Yang Liu

Towards Understanding the Fragility of Multilingual LLMs against Fine-Tuning Attacks (2025.findings-naacl)

Challenge:	Weakly supervised vision-and-language pre-training (WVLP) uses only local descriptions of images as cross-modal anchors to construct weakly-aligned image-text pairs for pre- training.
Approach:	They propose to take a small number of aligned image-text pairs as anchors and represent each unaligned image and text by its similarities to these anchors.
Outcome:	The proposed model reduces the cost of pre-training while maintaining decent performance on downstream tasks.

Copied to clipboard

Samuele Poppi, Zheng Xin Yong, Yifei He, Bobbie Chern, Han Zhao, Aobo Yang, Jianfeng Chi

Robustness in NLP

Challenge:	Recent advances in Large Language Models have sparked concerns about their safety.
Approach:	They propose a method to identify safety-related information in the model parameter space . they propose to use a few adversarially chosen examples to fine-tune LLMs .
Outcome:	The proposed method can break safety alignment in multilingual LLMs using a few examples . it also shows that the proposed method jailbreaks LLM models adapted to new languages .

Pi-SQL: Enhancing Text-to-SQL with Fine-Grained Guidance from Pivot Programming Languages (2025.findings-emnlp)

Copied to clipboard

Yongdong Chi, Hanqing Wang, Yun Chen, Yan Yang, Jian Yang, Zonghan Yang, Xiao Yan, Guanhua Chen

Programming Languages in NLP

Challenge:	Existing prompt-based methods craft meticulous text guidelines and examples to facilitate SQL generation, but their accuracy is hindered by the large semantic gap between the texts and the low-resource SQL programs.
Approach:	They propose to use Python as a pivot to bridge between natural language query and SQL program.
Outcome:	The proposed method improves the execution accuracy of the best-performing baseline by up to 3.20.

Missing Modality meets Meta Sampling (M3S): An Efficient Universal Approach for Multimodal Sentiment Analysis with Missing Modality (2022.aacl-main)

Copied to clipboard

Haozhe Chi, Minghua Yang, Junhao Zhu, Guanhong Wang, Gaoang Wang

Sentiment Analysis

Challenge:	Existing methods to address missing modalities often assume a particular modality is completely missing due to recording or transmission error.
Approach:	They propose a missing modality-based meta-sampling approach for multimodal sentiment analysis with missing modalities . they conduct experiments on IEMOCAP, SIMS and CMU-MOSI datasets .
Outcome:	The proposed method significantly improves on existing models with a mixture of missing modalities.

InfoXLM: An Information-Theoretic Framework for Cross-Lingual Language Model Pre-Training (2021.naacl-main)

Copied to clipboard

Zewen Chi, Li Dong, Furu Wei, Nan Yang, Saksham Singhal, Wenhui Wang, Xia Song, Xian-Ling Mao, Heyan Huang, Ming Zhou