Papers by Siyuan Wang

ALaRM: Align Language Models via Hierarchical Rewards Modeling (2024.findings-acl)

Challenge:	Recent studies have focused on enhancing reward models through data improvements, following the conventional training framework for reward models that directly optimizes the predicted rewards.
Approach:	They propose a hybrid alignment framework HAF-RM that incorporates additional constraint on token-level policy probabilities in addition to the reward score.
Outcome:	The proposed framework can supervise the internal preference model at the token level and optimize the mapping layer of the reward model at sequence level.

Copied to clipboard

Yuhang Lai, Siyuan Wang, Shujun Liu, Xuanjing Huang, Zhongyu Wei

Mirror-Consistency: Harnessing Inconsistency in Majority Voting (2024.findings-emnlp)

Challenge:	Current alignment approaches struggle with inconsistency and sparsity of human supervision signals.
Approach:	They propose a framework modeling hierarchical rewards in reinforcement learning from human feedback (RLHF) it integrates holistic rewards with aspect-specific rewards to enhance alignment of large language models with human preferences.
Outcome:	The proposed framework improves the alignment of large language models with human preferences by integrating holistic rewards with aspect-specific rewards.

Copied to clipboard

Siyuan Huang, Zhiyuan Ma, Jintao Du, Changhua Meng, Weiqiang Wang, Zhouhan Lin

Negative Sample is Negative in Its Own Way: Tailoring Negative Sentences for Image-Text Retrieval (2022.findings-naacl)

Challenge:	Large Language Models (LLMs) are a widely-used decoding strategy that relies on the plurality voting rule, which focuses on the most frequent answer while overlooking all other minority responses.
Approach:	They propose to incorporate a ‘reflective mirror’ into the self-ensemble decoding process and enables LLMs to critically examine inconsistencies among multiple generations.
Outcome:	The proposed method incorporates a ‘reflective mirror’ into the self-ensemble decoding process and enables LLMs to critically examine inconsistencies among multiple generations.

Copied to clipboard

Zhihao Fan, Zhongyu Wei, Zejun Li, Siyuan Wang, Xuanjing Huang, Jianqing Fan

Diagnosing Memorization in Chain-of-Thought Reasoning, One Token at a Time (2025.emnlp-main)

Challenge:	Existing approaches to retrieve hard negative sentences are limited in the scale of the dataset thus fail to identify negative samples of high difficulty for every image.
Approach:	They propose to use a model to generate synthetic negative sentences with higher difficulty by masking and refilling the images and performing word discrimination and word correction tasks to improve retrieval and generation.
Outcome:	The proposed model generates synthetic negative sentences with higher difficulty on MS-COCO and Flickr30K and is robust and faithful to state-of-the-art training.

Copied to clipboard

Huihan Li, You Chen, Siyuan Wang, Yixin He, Ninareh Mehrabi, Rahul Gupta, Xiang Ren

Self-supervised Semantic-driven Phoneme Discovery for Zero-resource Speech Recognition (2022.acl-long)

Challenge:	Large Language Models (LLMs) perform well on reasoning benchmarks but often fail when inputs alter slightly, raising concerns about overreliance on memorization.
Approach:	They propose a framework for Source-aware Token-level Identification of Memorization which attributes each token in a reasoning chain to one of multiple memorization sources based on their statistical co-occurrence with the token in the pretraining corpus.
Outcome:	The proposed framework attributes each token in a reasoning chain to one of multiple memorization sources based on their statistical co-occurrence with the token in the pretraining corpus.

Copied to clipboard

Liming Wang, Siyuan Feng, Mark Hasegawa-Johnson, Chang Yoo

Stepwise Informativeness Search for Improving LLM Reasoning (2025.emnlp-main)

Challenge:	a new approach for self-supervised speech representation learning is proposed . a phoneme inventory learning model is based on a discrete representation of speech .
Approach:	They propose a neural discrete representation learning model for self-supervised phoneme inventory learning with raw speech and word labels.
Outcome:	The proposed model learns better phoneme-level representations and lowers error rates on TIMIT and Mboshi benchmarks than previous state-of-the-art models.

Copied to clipboard

Siyuan Wang, Enda Zhao, Xiang Ren

Query Structure Modeling for Inductive Logical Reasoning Over Knowledge Graphs (2023.acl-long)

Challenge:	Recent advances in large language models have improved multistep reasoning but they lose focus over the middle of long contexts.
Approach:	They propose a tree search framework that proactively identifies underutilized steps and minimizing redundant information between steps.
Outcome:	The proposed framework generates more accurate and concise rationales with reduced errors and redundancy.

Copied to clipboard

Siyuan Wang, Zhongyu Wei, Meng Han, Zhihao Fan, Haijun Shan, Qi Zhang, Xuanjing Huang

Logic-Driven Context Extension and Data Augmentation for Logical Reasoning of Text (2022.findings-acl)

Challenge:	Existing methods for inductive reasoning over knowledge graphs lack the ability to model the logical structures of complex queries.
Approach:	They propose a structure-modeled textual encoding framework for inductive logical reasoning over KGs that encodes linearized query structures and entities using pre-trained language models to find answers.
Outcome:	The proposed framework encodes query structures and entities using pre-trained language models to find answers.

Copied to clipboard

Siyuan Wang, Wanjun Zhong, Duyu Tang, Zhongyu Wei, Zhihao Fan, Daxin Jiang, Ming Zhou, Nan Duan

Taming LLMs with Gradient Grouping (2025.acl-long)

Challenge:	Existing methods for logical reasoning of text focus on contextual semantics while struggling to explicitly model the logical inference process.
Approach:	They propose a logic-driven context extension framework and a data-driven augmentation algorithm that uses contrastive learning to better capture logical information.
Outcome:	The proposed framework outperforms existing methods on two benchmark datasets, ReClor and LogiQA.

Copied to clipboard

Siyuan Li, Juanxi Tian, Zedong Wang, Xin Jin, Zicheng Liu, Wentao Zhang, Dan Xu

From LLMs to MLLMs: Exploring the Landscape of Multimodal Jailbreaking (2024.emnlp-main)

Challenge:	a new study presents scaling with gradient grouping (SGG) the adaptive learning rate scaling approach is based on per-parameter statistics, which incurs memory overhead.
Approach:	They propose an optimizer wrapper that improves adaptive learning rate estimation by dynamic grouping and group-specific scaling.
Outcome:	The proposed algorithm improves learning rate estimation on diverse models with different model sizes and batch sizes.

Copied to clipboard

Siyuan Wang, Zhuohan Long, Zhihao Fan, Zhongyu Wei

Gumbel Reranking: Differentiable End-to-End Reranker Optimization (2025.acl-long)

Challenge:	Recent advances in Large Language Models (LLMs) have demonstrated remarkable performance across various tasks, effectively following instructions to meet diverse user needs.
Approach:	They propose a framework for evaluation benchmarks and attack techniques for LLMs and MLLMs to enhance their security.
Outcome:	The proposed frameworks have been exploited to exploit the weaknesses of LLMs and MLLMs.

Copied to clipboard

Siyuan Huang, Zhiyuan Ma, Jintao Du, Changhua Meng, Weiqiang Wang, Jingwen Leng, Minyi Guo, Zhouhan Lin

PathQG: Neural Question Generation from Facts (2020.emnlp-main)

Challenge:	Existing distillation-based approaches suffer from training-inference misalignment and fail to capture interdependencies among candidate documents.
Approach:	They propose a method to optimize rerankers by learning a stochastic, document-wise Top-k attention mask using the Gumbel Trick and Relaxed Top-K Sampling.
Outcome:	The proposed framework minimizes the overall language loss and improves recall on hotpotQA.

Copied to clipboard

Siyuan Wang, Zhongyu Wei, Zhihao Fan, Zengfeng Huang, Weijian Sun, Qi Zhang, Xuanjing Huang

Representation Learning

Challenge:	Existing research for question generation encodes text as a sequence of tokens without explicitly modeling fact information.
Approach:	They propose to incorporate facts in the input text for question generation in a comprehensive way.
Outcome:	The proposed model outperforms state-of-the-art models and human evaluation shows it generates relevant and informative questions.

SOLAR: Serendipity Optimized Language Model Aligned for Recommendation (2025.findings-emnlp)

Copied to clipboard

Zichen Yuan, Lifan Sun, Yucen Zhuang, Yue Wang, Xinyuan Song, Tianqi Xu, Siyuan Li, Junchen Fu, Youhua Li, Sirui Hong, Jiaqi Chen, Joemon M. Jose, Yongxin Ni

Multi-Objective Forward Reasoning and Multi-Reward Backward Refinement for Product Review Summarization (2024.lrec-main)

Challenge:	Large Language Models have shown strong potential in recommendation tasks . however, their application to serendipity-oriented recommendations remains challenging .
Approach:	They propose a domain-adaptive instruction tuning method that aligns Large Language Models with recommendation tasks.
Outcome:	The proposed framework bridges the domain gap between LLMs and recommendation tasks.

Copied to clipboard

Libo Sun, Siyuan Wang, Meng Han, Ruofei Lai, Xinyu Zhang, Xuanjing Huang, Zhongyu Wei

Text Generation

Challenge:	Product review summarization aims to generate a concise summary based on product reviews . factual accuracy, aspect comprehensiveness, and content relevance are challenges .
Approach:	They propose an FB-Thinker framework to improve product review summarization ability . they propose two Chinese product review summary datasets for instruction-tuning and evaluation .
Outcome:	The proposed framework improves product review summarization with forward reasoning and backward refinement.

Benchmark Self-Evolving: A Multi-Agent Framework for Dynamic LLM Evaluation (2025.coling-main)

Copied to clipboard

Siyuan Wang, Zhuohan Long, Zhihao Fan, Xuanjing Huang, Zhongyu Wei

Robustness in NLP

Challenge:	Recent advances in Large Language Models have demonstrated remarkable performance across tasks.
Approach:	They propose a benchmark self-evolving framework to dynamically evaluate rapidly advancing Large Language Models.
Outcome:	The proposed framework extends existing benchmarks to extend models across tasks and tasks.

Data Augmentation for Multiclass Utterance Classification – A Systematic Study (2020.coling-main)

Copied to clipboard

Binxia Xu, Siyuan Qiu, Jie Zhang, Yafang Wang, Xiaoyu Shen, Gerard de Melo

Information Extraction & Text Mining

Challenge:	a lack of sufficient training data for some categories can cause imbalanced data distributions . a weak classifier may miscategorize a request, resulting in customer dissatisfaction .
Approach:	They propose to use random resampling, word-level transformations and neural text generation to augment existing data to cope with imbalanced data.
Outcome:	The proposed methods improve utterance classification results by drawing on utterant variation.

Autoregressive Semantic Visual Reconstruction Helps VLMs Understand Better (2026.findings-acl)

Copied to clipboard

Dianyi Wang, Wei Song, Yikun Wang, Siyuan Wang, Kaicheng Yu, Zhongyu Wei, Jiaqi Wang

A Self-training Framework for Automated Medical Report Generation (2023.emnlp-main)

Challenge:	Typical large vision-language models emphasize vision-to-language alignment while overlooking fine-grained visual information.
Approach:	They introduce autoregressive semantic visual reconstruction (ASVR) that enables joint learning of visual and textual modalities within a unified autoregression framework.
Outcome:	The proposed model improves baselines and multimodal understanding benchmarks by 2-3%.

Copied to clipboard

Siyuan Wang, Zheng Liu, Bo Peng

Attribution-Based Analysis and Optimization of Modular Agentic Workflows (2026.findings-acl)

Challenge:	Medical report generation is an important medical artificial intelligence task.
Approach:	They propose a framework for medical report generation that exploits unlabeled medical images and a reference-free evaluation metric.
Outcome:	The proposed framework performs better than previous fully-supervised models trained on entire training data.

Copied to clipboard

Yingxuan Yang, Bo Huang, Siyuan Qi, Chao Feng, Haoyi Hu, Yuxuan Zhu, Jinbo Hu, Haoran Zhao, Ziyi He, Xiao Liu, ZongYu Wang, Muning Wen, Lin Qiu, Xuezhi Cao, Xunliang Cai, Yong Yu, Weinan Zhang

Challenge:	Large Language Models (LLMs) have driven the rise of agentic workflows . yet, how can we attribute performance gains to individual upgrades and their interactions?
Approach:	They propose a game-theoretic framework that models component upgrades as players and evaluates component coalitions to compute Shapley values.
Outcome:	The proposed framework provides interaction-aware attribution and recommendation for model allocation under a fixed workflow structure.

Multi-Agent Simulator Drives Language Models for Legal Intensive Interaction (2025.findings-naacl)

Copied to clipboard

ShengbinYue ShengbinYue, Ting Huang, Zheng Jia, Siyuan Wang, Shujun Liu, Yun Song, Xuanjing Huang, Zhongyu Wei

A Systematic Assessment of Language Models with Linguistic Minimal Pairs in Chinese (2026.tacl-1)

Challenge:	Large Language Models (LLMs) have advanced legal intelligence, but the scarcity of scenario data impedes the progress toward interactive legal scenarios.
Approach:	They propose a Multi-agent Legal Simulation Driver to generate synthetic data by simulating interactive legal scenarios.
Outcome:	The proposed framework ensures consistency of legal attributes between participants and introduces a supervisory mechanism to align participants’ characters and behaviors as well as addressing distractions.

Copied to clipboard

Yikang Liu, Yeting Shen, Hongao Zhu, Lilong Xu, Zhiheng Qian, Siyuan Song, Kejia Zhang, Jialong Tang, Pei Zhang, Baosong Yang, Rui Wang, Hai Hu

Mask Attention Networks: Rethinking and Strengthen Transformer (2021.naacl-main)

Challenge:	Using sub-linear length normalized log-probabilities (SLLN-LP), we find unequal lengths of sentences in minimal pairs difficult for LMs even up to 32B parameters.
Approach:	They propose to use ZhoBLiMP as a linguistic minimal pair benchmark for Chinese language models to mitigate biases.
Outcome:	The proposed metric mitigates biases in Chinese language models with over 100 paradigms . Anaphor, Quantifiers, and Ellipsis are difficult for LMs even up to 32B parameters .

Copied to clipboard

Zhihao Fan, Yeyun Gong, Dayiheng Liu, Zhongyu Wei, Siyuan Wang, Jian Jiao, Nan Duan, Ruofei Zhang, Xuanjing Huang

Training LLMs to be Better Text Embedders through Bidirectional Reconstruction (2025.emnlp-main)

Challenge:	Existing research explores to enhance the two sublayers separately to improve the capability of Transformer for text representation.
Approach:	They propose to combine SAN and Feed-Forward Networks to create a dynamic mask attention network with a learnable mask matrix which can model localness adaptively.
Outcome:	The proposed model outperforms the original Transformer on translation and text summarization tasks.

Copied to clipboard

Chang Su, Dengliang Shi, Siyuan Huang, Jintao Du, Changhua Meng, Yu Cheng, Weiqiang Wang, Zhouhan Lin

Ready Jurist One: Benchmarking Language Agents for Legal Intelligence in Dynamic Environments (2026.acl-long)

Challenge:	Existing text embedding approaches often leverage the embeddment of the final token, typically a reserved special token such as ‘[EOS]‘.
Approach:	They propose to add a new training stage before contrastive learning to enrich the semantics of the final token embedding.
Outcome:	The proposed training stage improves performance on the Massive Text Embedding Benchmark (MTEB), achieving new state-of-the-art results across different LLM base models and scales.

Copied to clipboard

Zheng Jia, Shengbin Yue, Wei Chen, Siyuan Wang, Yidong Liu, Zejun Li, Yun Song, Zhongyu Wei

Challenge:	Existing benchmarks for legal intelligence are limited to static evaluation paradigms or simplified scenarios.
Approach:	They introduce J1-ENVS, the first interactive and dynamic legal environment tailored for LLM-based agents.
Outcome:	The proposed framework assesses task performance and procedural compliance across legal proficiency levels.

Guardian-as-an-Advisor: Advancing Next-Generation Guardian Models for Trustworthy LLMs (2026.findings-acl)

Copied to clipboard

Yue Huang, Haomin Zhuang, Jiayi Ye, Han Bao, Yanbo Wang, Hang Hua, Siyuan Wu, Pin-Yu Chen, Xiangliang Zhang

Can We Edit Multimodal Large Language Models? (2023.emnlp-main)

Challenge:	prevailing taxonomies neglect robustness and honesty, yielding safer-on-paper but less useful systems.
Approach:	They propose a soft-gating pipeline where a guardian predicts a binary risk label plus a concise explanation and prepends this advice to the original query for re-inference.
Outcome:	The proposed model maintains safety while reducing over-refusal.

Copied to clipboard

Siyuan Cheng, Bozhong Tian, Qingbin Liu, Xi Chen, Yongheng Wang, Huajun Chen, Ningyu Zhang

Editing Large Language Models: Problems, Methods, and Opportunities (2023.emnlp-main)

Challenge:	Existing methods to edit multimodal models have been used to incrementally infuse a language model with a new set of facts.
Approach:	They construct a benchmark for editing multimodal Large Language Models and establish metrics for evaluation.
Outcome:	The proposed benchmarks show that editing multimodal models is not as difficult as editing single-modal models.

Copied to clipboard

Yunzhi Yao, Peng Wang, Bozhong Tian, Siyuan Cheng, Zhoubo Li, Shumin Deng, Huajun Chen, Ningyu Zhang

A Structure-Aware Argument Encoder for Literature Discourse Analysis (2022.coling-1)

Challenge:	Recent advances in model editing for LLMs have created challenges and opportunities for the community.
Approach:	They propose to alter the behavior of LLMs efficiently within a specific domain without negatively impacting performance across other inputs.
Outcome:	The proposed method alters behavior of LLMs efficiently within a specific domain without negatively impacting performance across other inputs.

Copied to clipboard

Yinzi Li, Wei Chen, Zhongyu Wei, Yujun Huang, Chujun Wang, Siyuan Wang, Qi Zhang, Xuanjing Huang, Libo Wu

AMEX: Android Multi-annotation Expo Dataset for Mobile GUI Agents (2025.findings-acl)

Challenge:	Existing research for argument representation learning treats tokens in sentences equally and ignores the implied structure information of argumentative context.
Approach:	They propose to separate tokens into two groups to capture structural information of arguments and to incorporate paragraph-level position information into the model.
Outcome:	The proposed model captures structural information of arguments and is able to identify arguments automatically.

Copied to clipboard

Yuxiang Chai, Siyuan Huang, Yazhe Niu, Han Xiao, Liang Liu, Guozhi Wang, Dingyu Zhang, Shuai Ren, Hongsheng Li

Challenge:	a new dataset is being developed to improve the capabilities of mobile GUI-control agents.
Approach:	They propose a dataset designed for generalist mobile GUI-control agents . they use screenshots from popular mobile applications to create a detailed GUI-annotated dataset .
Outcome:	The Android Multi-annotation EXpo (AMEX) is a large-scale dataset for generalist mobile GUI-control agents . it includes screenshots from popular mobile applications, which are annotated at multiple levels .

SparkRA: A Retrieval-Augmented Knowledge Service System Based on Spark Large Language Model (2024.emnlp-demo)

Copied to clipboard

Dayong Wu, Jiaqi Li, Baoxin Wang, Honghong Zhao, Siyuan Xue, Yanjie Yang, Zhijun Chang, Rui Zhang, Li Qian, Bo Wang, Shijin Wang, Zhixiong Zhang, Guoping Hu

A Reinforcement Learning Framework for Natural Question Generation using Bi-discriminators (C18-1)

Challenge:	Large language models (LLMs) have shown remarkable achievements across various language tasks.
Approach:	They propose a scientific literature LLM and a knowledge service system based on it . they collect scientific literature and then pre-train it using autoregressive training .
Outcome:	The proposed system provides literature investigation, paper reading, and academic writing functions.

Copied to clipboard

Zhihao Fan, Zhongyu Wei, Siyuan Wang, Yang Liu, Xuanjing Huang

Text Generation

Challenge:	Existing research on visual question generation is focused on training models to fit the annotated data set that makes them indifferent from other language generation tasks.
Approach:	They propose to use two discriminators to enhance the training of a visual question generator to ask natural questions about an image.
Outcome:	The proposed model outperforms state-of-the-art models in terms of automatic and human evaluation metrics.

Simple-VGC: Enhancing Visual Grounding in Multimodal Reasoning via Adaptive Tool Composition (2026.acl-long)

Copied to clipboard

Ye Wang, Qianglong Chen, Siyuan Wang, Zejun Li, Shijie Guo, Zhirui Zhang, Zhongyu Wei

A3: Android Agent Arena for Mobile GUI Agents with Essential-State Procedural Evaluation (2026.findings-acl)

Challenge:	Existing multimodal large language models suffer from systematic failures in basic visual understanding.
Approach:	They propose a tool-augmented reasoning framework with three targeted compensation strategies to address these problems.
Outcome:	The proposed framework improves visual grounding by re-injecting the original image to mitigate visual forgetting, the authors show . the proposed framework also improves the accuracy of the visual inputs, the researchers show - and the results are promising .

Copied to clipboard

Yuxiang Chai, Shunye Tang, Han Xiao, Weifeng Lin, Hanhao Li, Jiayu Zhang, Liang Liu, Pengxiang Zhao, Guangyi Liu, Guozhi Wang, Shuai Ren, Rongduo Han, Haining Zhang, Siyuan Huang, Hongsheng Li

AutoJudger: An Agent-Driven Framework for Efficient Benchmarking of MLLMs (2026.acl-long)

Challenge:	Existing evaluation methods for mobile GUI agents rely on static frame assessments or offline static apps.
Approach:	They propose an evaluation system that leverages large language models as reward models to verify task completion and process achievement.
Outcome:	The proposed system addresses the limitations of traditional function based evaluation methods on online dynamic apps.

Copied to clipboard

Xuanwen Ding, Chengjun Pan, Zejun Li, Jiwen Zhang, Siyuan Wang, Zhongyu Wei

Green & Sustainable NLP

Challenge:	Evaluating multimodal large language models (MLLMs) is becoming increasingly expensive as benchmarks grow in scale and cross-modality complexity.
Approach:	They propose an adaptive evaluation framework for efficient benchmarking that treats evaluation as an interview-like process by keeping a hypothesized ability structure of the evaluated model and actively selecting the informative questions.
Outcome:	Experiments on four representative multimodal benchmarks show that **A2-Judger significantly improves sample efficiency while maintaining reliable evaluation results.

View Dialogue in 2D: A Two-stream Model in Time-speaker Perspective for Dialogue Summarization and beyond (2022.coling-1)

Copied to clipboard

Keli Xie, Dongchen He, Jiaxin Zhuang, Siyuan Lu, Zhongfeng Wang

Natural Language Interfaces

Challenge:	Existing models for dialogue summarization focus on document summarizing on time and speaker-centered points, but this approach is limited in understanding the dialogue.
Approach:	They propose a 2D view of dialogue based on a time-speaker perspective where the time and speaker streams of dialogue can be obtained as strengthened input.
Outcome:	The proposed model outperforms existing models on the QMSum dataset and improves summary faithfulness and human evaluation.

Bridging by Word: Image Grounded Vocabulary Construction for Visual Captioning (P19-1)

Copied to clipboard

Zhihao Fan, Zhongyu Wei, Siyuan Wang, Xuanjing Huang

PreAct: Prediction Enhances Agent’s Planning Ability (2025.coling-main)

Challenge:	Existing research on image captioning generates frequent n-grams with irrelevant words.
Approach:	They propose to construct an image-grounded vocabulary incorporating visual information and relations among words into the decoding process directly.
Outcome:	The proposed framework is compared with state-of-the-art models on MS COCO and Flickr30k and shows that it is more efficient than existing models.

Copied to clipboard

Dayuan Fu, Jianzhao Huang, Siyuan Lu, Guanting Dong, Yejie Wang, Keqing He, Weiran Xu

Symbolic Working Memory Enhances Language Models for Complex Rule Application (2024.emnlp-main)

Challenge:	Existing methods to analyze Markov decision processes (MDPs) are based on chain-of-thought (COT) and historical thought, action, and observation.
Approach:	They propose a model that integrates prediction, reasoning, and action with other models to provide a wider range of reasoning and more efficient actions.
Outcome:	The proposed model outperforms the ReAct method in completing complex tasks and is more efficient when paired with other memory or selection strategy techniques.

Copied to clipboard

Siyuan Wang, Zhongyu Wei, Yejin Choi, Xiang Ren

Interleaved Latent Visual Reasoning with Selective Perceptual Modeling (2026.acl-long)

Challenge:	Large Language Models (LLMs) excel in single-step rule application but struggle with multi-step deductive reasoning when rules are presented non-sequentially.
Approach:	They propose to augment LLMs with external working memory and introduce a neurosymbolic framework for rule application that stores facts and rules in both natural language and symbolic forms, enabling precise tracking.
Outcome:	The proposed framework iteratively performs symbolic rule grounding and LLM-based rule implementation.

Copied to clipboard

Shuai Dong, Siyuan Wang, Xingyu Liu, Chenglin Li, Haowen Hou, Zhongyu Wei

Thinking-Based Non-Thinking: Solving the Reward Hacking Problem in Training Hybrid Reasoning Models via Reinforcement Learning (2026.acl-long)

Challenge:	Existing approaches to interleaved reasoning are limited by the cost of re-encoding pixel-dense images.
Approach:	They propose a framework that unifies dynamic state evolution with precise perceptual modeling.
Outcome:	The proposed framework outperforms existing approaches on multimodal reasoning benchmarks.

Copied to clipboard

Siyuan Gan, Jiaheng Liu, Boyan Wang, Tianpei Yang, Runqing Miao, Yuyao Zhang, Fanyu Meng, Junlan Feng, Linjian Meng, Jing Huo, Yang Gao

EasyEdit: An Easy-to-use Knowledge Editing Framework for Large Language Models (2024.acl-demos)

Challenge:	Existing work on large reasoning models (LRMs) focuses on using reinforcement learning (RL) to train hybrid reasoning models that automatically decide whether to engage in thinking or not based on the complexity of the query.
Approach:	They propose to use reinforcement learning to train hybrid reasoning models that automatically decide whether to engage in thinking or not based on the complexity of the query.
Outcome:	The proposed model reduces token usage by around 50%$ compared to DeepSeek-R1-Distill-Qwen-1.5B/7B and DeepScaleR-1.5b, while significantly improving accuracy.

Copied to clipboard

Peng Wang, Ningyu Zhang, Bozhong Tian, Zekun Xi, Yunzhi Yao, Ziwen Xu, Mengru Wang, Shengyu Mao, Xiaohan Wang, Siyuan Cheng, Kangwei Liu, Yuansheng Ni, Guozhou Zheng, Huajun Chen

Locate Then Ask: Interpretable Stepwise Reasoning for Multi-hop Question Answering (2022.coling-1)

Challenge:	Large Language Models (LLMs) suffer from knowledge cutoff or fallacy issues, which means they are unaware of unseen events or generate text with incorrect facts owing to outdated/noisy data.
Approach:	They propose an easy-to-use knowledge editing framework for Large Language Models that allows users to easily edit updated knowledge and adjust undesired behavior while minimizing the impact on unrelated inputs.
Outcome:	The proposed framework surpasses traditional fine-tuning in terms of reliability and generalization.

Copied to clipboard

Siyuan Wang, Zhongyu Wei, Zhihao Fan, Qi Zhang, Xuanjing Huang

AI Hospital: Benchmarking Large Language Models in a Multi-agent Medical Interaction Simulator (2025.coling-main)

Challenge:	Existing methods for multi-hop reasoning ignore grounding on supporting facts of each step, which tends to generate inaccurate decompositions.
Approach:	They propose an interpretable stepwise reasoning framework that incorporates supporting sentences and questions at each intermediate step and utilizes the inference of the current hop for the next until reasoning out the final result.
Outcome:	The proposed model can boost performance and yield a better interpretable reasoning process without decomposition supervision.

Copied to clipboard

Zhihao Fan, Lai Wei, Jialong Tang, Wei Chen, Wang Siyuan, Zhongyu Wei, Fei Huang

MAGNET: Towards Adaptive GUI Agents with Memory-Driven Knowledge Evolution (2026.acl-long)

Challenge:	Recent large language models (LLMs) have demonstrated superior performance in static medical question answering benchmarks, rivaling even human experts.
Approach:	They propose a multi-agent framework emulating dynamic medical interactions between Doctor as player and NPCs including Patient and Examiner to assess the performance of LLM-driven Doctor agents in simulated clinical scenarios.
Outcome:	The proposed framework emulates dynamic medical interactions between Doctor as player and NPCs including Patient and Examiner.

Copied to clipboard

Libo Sun, Jiwen Zhang, Siyuan Wang, Zhongyu Wei

Challenge:	Mobile GUI agents powered by large foundation models can perform tasks autonomously, but frequent updates that alter UI appearance and reorganize workflows cause agents trained on historical data to fail.
Approach:	They propose a memory-driven adaptive agent framework with stationary memory that links visual features to stable functional semantics and procedural memory that captures stable task intents across varying workflows.
Outcome:	The proposed framework improves performance over memory-augmented baselines and offline benchmarks on AndroidWorld.

WISE: Weak-Supervision-Guided Step-by-Step Explanations for Multimodal LLMs in Image Classification (2025.emnlp-main)

Copied to clipboard

Yiwen Jiang, Deval Mehta, Siyuan Yan, Yaling Shen, Zimu Wang, Zongyuan Ge

An Enhanced Knowledge Injection Model for Commonsense Generation (2020.coling-main)

Challenge:	Existing MCoT methods focus on inter-object reasoning, overlooking intra-object understanding crucial for image classification.
Approach:	They propose a Weak-supervision-guided Step-by-step Explanation method that reformulates MCoTs under weak supervision into concise, interpretable reasoning chains.
Outcome:	The proposed method improves interpretability by 37% and improves classification accuracy.

Copied to clipboard

Zhihao Fan, Yeyun Gong, Zhongyu Wei, Siyuan Wang, Yameng Huang, Jian Jiao, Xuanjing Huang, Nan Duan, Ruofei Zhang

Activating Distributed Visual Region within LLMs for Efficient and Effective Vision-Language Training and Inference (2025.acl-long)

Challenge:	a recent study shows that digging the relationship of concepts from scratch is non-trivial for commonsense generation tasks.
Approach:	They use a retrieve-and-edit framework to retrieve a prototype with these concepts . they use qt and qq to generate commonsense questions at scale .
Outcome:	The proposed method significantly improves the performance on commonsense generation tasks.

Copied to clipboard

Siyuan Wang, Dianyi Wang, Chengxing Zhou, Zejun Li, Zhihao Fan, Xuanjing Huang, Zhongyu Wei

Visual Data in NLP

Challenge:	Existing Large Vision-Language Models (LVLMs) learn visual capacity through visual instruction tuning.
Approach:	They propose a method for LVLMs to be trained by selective layers tuning . they propose removing non-critical layers outside the visual region .
Outcome:	The proposed approach preserves nearly 99% of visual performance and improves textual task results while reducing training time.

Boosting LLM Agents with Recursive Contemplation for Effective Deception Handling (2024.findings-acl)

Copied to clipboard

Shenzhi Wang, Chang Liu, Zilong Zheng, Siyuan Qi, Shuo Chen, Qisen Yang, Andrew Zhao, Chaofei Wang, Shiji Song, Gao Huang

Challenge:	Recent advances in large language models (LLMs) have led to significant success in using LLMs as agents.
Approach:	They propose a cognitive framework that incorporates first-order and second-order perspective transitions into LLMs to enhance their ability to identify and counteract deceptive information.
Outcome:	The proposed framework enhances LLMs’ ability to identify and counteract deceptive information without extra fine-tuning and data.

Analytical Reasoning of Text (2022.findings-naacl)

Copied to clipboard

Wanjun Zhong, Siyuan Wang, Duyu Tang, Zenan Xu, Daya Guo, Yining Chen, Jiahai Wang, Jian Yin, Ming Zhou, Nan Duan

Fine-grained Medical Vision-Language Representation Learning for Radiology Report Generation (2023.emnlp-main)

Challenge:	Existing models with implicit reasoning ability struggle to solve analytical reasoning of text.
Approach:	They propose an approach to analyze text and use it to perform reasoning over it.
Outcome:	The proposed approach outperforms pre-trained models on an analysis of the Law School Admission Test dataset.

Copied to clipboard

Siyuan Wang, Bo Peng, Yichao Liu, Qi Peng

Visual Data in NLP

Challenge:	Existing methods to learn medical vision-language representations by contrasting images with entire reports are not effective.
Approach:	They propose a phenotype-driven medical vision-language representation learning framework to bridge the gap between visual and textual modalities for improved text-oriented generation.
Outcome:	The proposed framework bridges the gap between visual and textual modalities for improved radiology report generation.

How Jailbreak Defenses Work and Ensemble? A Mechanistic Investigation (2025.findings-emnlp)

Copied to clipboard

Zhuohan Long, Siyuan Wang, Shujun Liu, Yuhang Lai

Weights-Rotated Preference Optimization for Large Language Models (2025.emnlp-main)

Challenge:	Jailbreak attacks, where harmful prompts bypass generative models’ built-in safety, raise serious concerns about model vulnerability.
Approach:	They propose to reframe the standard generation task as a binary classification problem to assess model refusal tendencies for both harmful and benign queries.
Outcome:	The proposed defenses improve model safety or optimize the trade-off between safety and helpfulness.

Copied to clipboard

Chenxu Yang, Ruipeng Jia, Mingyu Zheng, Naibin Gu, Zheng Lin, Siyuan Chen, Weichong Yin, Hua Wu, Weiping Wang

In Search of the Long-Tail: Systematic Generation of Long-Tail Inferential Knowledge via Logical Rule Guided Search (2024.emnlp-main)

Challenge:	Existing methods to align large language models with high reward hacking are limited by the complexity of the parameter space and the complexity.
Approach:	They propose a weights-rotated preference optimization algorithm that constrains the output layer logits with the KL divergence inherited from DPO and fine-tunes the intermediate hidden states.
Outcome:	The proposed algorithm achieves a 3.27-point improvement on AlpacaEval 2 and surpasses the best baseline by 6.2 to 7.5 points on MT-Bench with merely 0.015% of the trainable parameters.

Copied to clipboard

Huihan Li, Yuting Ning, Zeyi Liao, Siyuan Wang, Xiang Li, Ximing Lu, Wenting Zhao, Faeze Brahman, Yejin Choi, Xiang Ren

Information Retrieval

Challenge:	Logic-Induced-Knowledge-Search (LINK) is a framework for generating factually-correct yet long-tail inferential knowledge.
Approach:	They introduce a framework to obtain factually-correct yet long-tail inferential statements using variable-wise prompting grounded on symbolic rules.
Outcome:	The proposed framework is able to obtain factually-correct yet long-tail inferential statements while ensuring factual correctness.

Can LLMs Reason with Rules? Logic Scaffolding for Stress-Testing and Improving LLMs (2024.acl-long)

Copied to clipboard

Siyuan Wang, Zhongyu Wei, Yejin Choi, Xiang Ren