LLMs and Natural Language Understanding

We recently work on LLMs, natural language understanding for solving math word problems, document summarization and sentimental analysis about Covid-19.

• Yue Huang, Chujie Gao, Siyuan Wu, Haoran Wang, Xiangqi Wang, Jiayi Ye, Yujun Zhou, Yanbo Wang, Jiawen Shi, Qihui Zhang, Han Bao, Zhaoyi Liu, Yuan Li, Tianrui Guan, Peiran Wang, Haomin Zhuang, Dongping Chen, Kehan Guo, Andy Zou, Bryan Hooi, Caiming Xiong, Elias Stengel-Eskin, Hongyang Zhang, Hongzhi Yin, Huan Zhang, Huaxiu Yao, Jieyu Zhang, Jaehong Yoon, Kai Shu, Ranjay Krishna, Swabha Swayamdipta, Weijia Shi, Xiang Li, Yuexing Hao, Zhihao Jia, Zhize Li, Xiuying Chen, Zhengzhong Tu, Xiyang Hu, Tianyi Zhou, Jieyu Zhao, Lichao Sun, Furong Huang, Or Cohen-Sasson, Prasanna Sattigeri, Anka Reuel, Max Lamparth, Yue Zhao, Nouha Dziri, Yu Su, Huan Sun, Heng Ji, Chaowei Xiao, Mohit Bansal, Nitesh V Chawla, Jian Pei, Jianfeng Gao, Michael Backes, Philip S. Yu, Neil Zhenqiang Gong, Pin-Yu Chen, Bo Li, Dawn Song, Xiangliang Zhang. TrustGen: A Platform of Dynamic Benchmarking on the Trustworthiness of Generative Foundation Models. Accepted by ICLR 2026.  (Acceptance rate ~ 28%, out of 19000 submissions) Dataset available.

• Yue Huang ~Yue_Huang9 , Hang Hua, Yujun Zhou, Pengcheng Jing, Manish Nagireddy, Inkit Padhi, Greta Dolcetti, Zhangchen Xu, Subhajit Chaudhury, Ambrish Rawat, Liubov Nedoshivina, Pin-Yu Chen, Prasanna Sattigeri, Xiangliang Zhang. Building a Foundational Guardrail for General Agentic Systems via Synthetic Data. Accepted by ICLR 2026.  (Acceptance rate ~ 28%, out of 19000 submissions) 

• Dawei Li, Renliang Sun, Yue Huang, Ming Zhong, Bohan Jiang, Jiawei Han, Xiangliang Zhang, Wei Wang, huan liu. Preference Leakage: A Contamination Problem in LLM-as-a-judge. Accepted by ICLR 2026.  (Acceptance rate ~ 28%, out of 19000 submissions) 

• Xingjian Hu, Ziqian Zhang, Yue Huang, Kai Zhang, Ruoxi Chen, Yixin Liu, Qingsong Wen, Kaidi Xu, Xiangliang Zhang, Neil Zhenqiang Gong, Lichao Sun. RankLLM: Weighted Ranking of LLMs by Quantifying Question Difficulty. Accepted by ICLR 2026.  (Acceptance rate ~ 28%, out of 19000 submissions) 

• Xiaobo Xing, Wei Yuan, Tong Chen, Quoc Viet Hung Nguyen, Xiangliang Zhang, Hongzhi Yin. TableDART: Dynamic Adaptive Multi-Modal Routing for Table Understanding. Accepted by ICLR 2026.  (Acceptance rate ~ 28%, out of 19000 submissions) 

• Xiaonan Luo, Yue Huang, Ping He, Xiangliang Zhang. Better Datasets Start From RefineLab: Automatic Optimization for High-Quality Dataset Refinement. Accepted by AAAI 2026. (4,167 out of 23,680 submissions, acceptance rate 17.6%)

• Yanchi Ru, Yue Huang, Xiangliang Zhang. RMO: Towards Better LLM Alignment via Reshaping Reward Margin Distributions. Accepted by AAAI 2026. (4,167 out of 23,680 submissions, acceptance rate 17.6%)

• Yue Huang, Xiangqi Wang, Xiangliang Zhang. SPA: Achieving Consensus in LLM Alignment via Self-Priority Optimization. Accepted by AAAI 2026. (4,167 out of 23,680 submissions, acceptance rate 17.6%)

• Yujun Zhou, Jingdong Yang, Yue Huang, Kehan Guo, Zoe Emory, Bikram Ghosh, Amita Bedar, Sujay Shekar, Zhenwen Liang, Pin-Yu Chen, Tian Gao, Werner Geyer, Nuno Moniz, Nitesh V. Chawla, and Xiangliang Zhang. Benchmarking Large Language Models on Safety Issues in Scientific Labs. Nature Machine Intelligence, Jan 2026. https://doi.org/10.1038/s42256-025-01152-1 Highlighted by New Scientist magazine, and Science News.

• Yue Huang, Zhengzhe Jiang, Xiaonan Luo, Kehan Guo, Haomin Zhuang, Yujun Zhou, Zhengqing Yuan, Xiaoqi Sun, Jules Schleinitz, Yanbo Wang, Shuhao Zhang, Mihir Surve, Nitesh V Chawla, Olaf Wiest, Xiangliang Zhang. ChemOrch: Empowering LLMs with Chemical Intelligence via Groundbreaking Synthetic Instructions. Accepted by NeurIPS 2025 (Accepted 5290 (24.52%) papers  out of 21575 submissions).  

• Xiangqi Wang, Yue Huang, Yanbo Wang, Xiaonan Luo, Kehan Guo, Yujun Zhou, Xiangliang Zhang. AdaReasoner: Adaptive Reasoning Enables More Flexible Thinking. Accepted by NeurIPS 2025 (Accepted 5290 (24.52%) papers out of 21575 submissions).  

• Yanbo Wang, Zixiang Xu, Yue Huang, Xiangqi Wang, Zirui Song, Lang Gao, Chenxi Wang, Xiangru Tang, Yue Zhao, Arman Cohan, Xiangliang Zhang, Xiuying Chen.DyFlow: Dynamic Workflow Framework for Agentic Reasoning. Accepted by NeurIPS 2025 (Accepted 5290 (24.52%) papers out of 21575 submissions).  

• Yanbo Wang, Zixiang Xu, Yue Huang, Chujie Gao, Siyuan Wu, Jiayi Ye, Pin-Yu Chen, Xiuying Chen, Xiangliang Zhang. Adaptive Distraction: Probing LLM Contextual Robustness with Automated Tree Search. Accepted by NeurIPS 2025 (Accepted 5290 (24.52%) papers out of 21575 submissions).  

• Anna Sokol, Elizabeth M. Daly, Michael Hind, David Piorkowski, Xiangliang Zhang, Nuno Moniz, Nitesh V Chawla. BenchmarkCards: Standardized Documentation for Large Language Model Benchmarks. Accepted by NeurIPS 2025 Datasets and Benchmarks Track (Accepted 497 (24.91%) papers out of 1995 submissions).      

• Yue Huang, Zhengqing Yuan, Yujun Zhou, Kehan Guo, Xiangqi Wang, Haomin Zhuang, Weixiang Sun, Lichao Sun, Jindong Wang, Yanfang Ye, Xiangliang Zhang. Exposing and Patching the Flaws of Large Language Models in Social Character Simulation. Accepted by COLM 2025. (acceptance rate 32%, 418 out of 1,305 submissions)

• Yanbo Wang, Jiayi Ye, Siyuan Wu, Chujie Gao, Yue Huang, Xiuying Chen, Yue Zhao, Xiangliang Zhang. TrustEval: A Dynamic Evaluation Toolkit on Trustworthiness of Generative Foundation Models. NAACL 2025 (System Demonstrations).

• Yue Huang, Siyuan Wu, Chujie Gao, Dongping Chen, Qihui Zhang, Yao Wan, Tianyi Zhou, Chaowei Xiao, Jianfeng Gao, Xiangliang Zhang, Lichao Sun. DataGen: Unified Synthetic Dataset Generation via Large Language Models. Accepted by ICLR 2025.

• Jiayi Ye, Yanbo Wang, Yue Huang, Dongping Chen, Qihui Zhang, Nuno Moniz, Tian Gao, Werner Geyer, Chao Huang, Pin-Yu Chen, Nitesh V Chawla, Xiangliang Zhang. Justice or Prejudice? Quantifying Biases in LLM-as-a-Judge. Accepted by ICLR 2025.

• Yi Gui, Yao Wan, Zhen Li, Zhongyi Zhang, Dongping Chen, Hongyu Zhang, Yi Su, Bohua Chen, Xing Zhou, Wenbin Jiang, Xiangliang Zhang. UICopilot: Automating UI Synthesis via Hierarchical Code Generation from Webpage Designs. Accepted by TheWebConf 2025.

• Yi Gui, Zhen Li, Yao Wan, Yemin Shi, Hongyu Zhang, Yi Su, Bohua Chen, Dongping Chen, Siyuan Wu, Xing Zhou, Wenbin Jiang, Hai Jin, Xiangliang Zhang. WebCode2M: A Real-World Dataset for Code Generation from Webpage Designs. Accepted by TheWebConf 2025.

• Yue Huang, Chenrui Fan, Yuan Li, Siyuan Wu, Tianyi Zhou, Xiangliang Zhang, Lichao Sun. 1+1>2: Can Large Language Models Serve as Cross-Lingual Knowledge Aggregators? Accepted by EMNLP 2024 Main. arXiv link.

• Yujun Zhou, Yufei Han, Haomin Zhuang, Kehan Guo, Zhenwen Liang, Hongyan Bao, Xiangliang Zhang. Defending Jailbreak Prompts via In-Context Adversarial Game. Accepted by EMNLP 2024 Main. arXiv link.

• Ziyi Kou, Shichao Pei, Meng Jiang, Xiangliang Zhang. RAt: Injecting Implicit Bias for Text-To-Image Prompt Refinement Models. Accepted by EMNLP 2024 Main.

• Tianyu Yang, Yiyang Nan, Lisen Dai, Zhenwen Liang, Yapeng Tian, Xiangliang Zhang. SaSR-Net: Source-Aware Semantic Representation Network for Enhancing Audio-Visual Question Answering. Accepted by EMNLP 2024 Findings

• Xiuying Chen, Mingzhe Li, Shen Gao, Xin Cheng, Qingqing Zhu, Rui Yan, Xin Gao, Xiangliang Zhang. Flexible and Adaptable Summarization via Expertise Separation. Accepted by SIGIR 2024, 14-18 July, 2024 in Washington D.C., USA. (20.1% acceptance rate, 159 out of 791 submissions).

• Zhenwen Liang, Kehan Guo, Gang Liu, Taicheng Guo, Yujun Zhou, Tianyu Yang, Jiajun Jiao, Renjie Pi, Jipeng Zhang, Xiangliang Zhang: SceMQA: A Scientific College Entrance Level Multimodal Question Answering Benchmark. To appear in ACL 2024. Bangkok, Thailand from August 11th to 16th, 2024.

• Xiuying Chen, Shen Gao, Mingzhe Li, Qingqing Zhu, Xin Gao, Xiangliang Zhang: Write Summary Step-by-Step: A Pilot Study of Stepwise Summarization. IEEE ACM Transactions Audio Speech Language Process. 32: 1406-1415 (2024)

• Zhenwen Liang, Wenhao Yu, Tanmay Rajpurohit, Peter Clark, Xiangliang Zhang, Ashwin Kalyan. Let GPT be a Math Tutor: Teaching Math Word Problem Solvers with Customized Exercise Generation. Accepted by EMNLP 2023, Singapore from Dec 6th to Dec 10th, 2023. Paper on arXiv

• Zhenwen Liang, Tianyu Yang, Jipeng Zhang, Xiangliang Zhang. UniMath: A Foundational and Multimodal Mathematical Reasoner. Accepted by EMNLP 2023, Singapore from Dec 6th to Dec 10th, 2023.

• Xiuying Chen, Guodong Long, Chongyang Tao, Mingzhe Li, Xin Gao, Chengqi Zhang, Xiangliang Zhang. Improving the Robustness of Summarization Systems with Dual Augmentation. Accepted by ACL 2023. Toronto, Canada. July 9-14, 2023.

• Zhenwen Liang, Jipeng Zhang, Kehan Guo, Xiaodong Wu, Jie Shao, Xiangliang Zhang. Compositional Mathematical Encoding for Math Word Problems. Accepted by the Findings of ACL 2023.

• Xiuying Chen, Mingzhe Li, Shen Gao, Xin Cheng, Qiang Yang, Qishen Zhang, Xin Gao and Xiangliang Zhang. A Topic-aware Summarization Framework with Different Modal Side Information. Accepted by SIGIR 2023 (Full paper, Acceptance rate = 165/822 = 20.1%)

• Zhenwen Liang, Jipeng ZHANG,  Lei Wang, Yan Wang, Jie Shao, Xiangliang Zhang. Generalizing Math Word Problem Solvers via Solution Diversification. Accepted by the Thirty-Seventh AAAI Conference on Artificial Intelligence (AAAI 2023). Feb 7-14, 2023 Washington DC. (Acceptance rate = 19.6% （1,721 of 8,777 submissions))

• Xiuying Chen, Mingzhe Li, Jiayi Zhang, Xiaoqiang Xia, Chen Wei, Jianwei Cui, Xin Gao, Xiangliang Zhang, Rui Yan. Learning towards Selective Data Augmentation for Dialogue Generation. Accepted by the Thirty-Seventh AAAI Conference on Artificial Intelligence (AAAI 2023). Feb 7-14, 2023 Washington DC. (Acceptance rate = 19.6% （1,721 of 8,777 submissions))

• Zhenwen Liang, Jipeng Zhang and Xiangliang Zhang Analogical Math Word Problems Solving with Enhanced Problem-Solution Association. Accepted by EMNLP 2022. 

• Xiuying Chen, Mingzhe Li, Shen Gao, Rui Yan, Xin Gao and Xiangliang Zhang. Scientific Paper Extractive Summarization Enhanced by Citation Graphs. Accepted by EMNLP 2022. 

• Youssef Sherif Mansour Mohamed, Shyma Yaser Alhuwaider, Mohamed Abdelfattah, Feifan Li, Kenneth Ward Church, Xiangliang Zhang and Mohamed Elhoseiny. ArtELingo: A Million Emotion Annotations of WikiArt with Emphasis on Diversity over Language and Culture. Accepted by EMNLP 2022. 

• Xiuying Chen, Mingzhe Li, Xin Gao, Xiangliang Zhang. Towards Improving Faithfulness in Abstractive Summarization. Accepted by NeurIPS 2022.

• Xiuying Chen, Hind Alamro, Mingzhe Li, Shen Gao, Rui Yan, Xin Gao and Xiangliang Zhang. Target-aware Abstractive Related Work Generation with Contrastive Learning. Accepted by SIGIR 2022. 

• Zhenwen Liang, Jipeng Zhang, Lei Wang, Wei QIN, Yunshi Lan, Jie Shao, Xiangliang Zhang. MWP-BERT: Numeracy-Augmented Pre-training for Math Word Problem Solving. To appear in the Findings of NAACL 2022. Early Access on arXiv

• Reem Alghamdi, Zhenwen Liang, Xiangliang Zhang. ArMATH: a Dataset for Solving Arabic Math Word Problems. Accepted by the 13th International Conference on Language Resources and Evaluation (LREC 2022). The dataset is available at GitHub

• Xiuying Chen, Mingzhe Li, Shen Gao, Zhangming Chan, Dongyan Zhao, Xin Gao, Xiangliang Zhang, Rui Yan. Follow the Timeline! Generating Abstractive and Extractive Timeline Summary in Chronological Order. ACM Transactions on Information Systems (TOIS). February 2022. https://doi.org/10.1145/3517221

• Zhenwen Liang, Xiangliang Zhang. Data-Efficient Language Shaped Few-shot Image Classification. Accepted in Findings of EMNLP 2021.

• Xiuying Chen, Hind Alamro, Mingzhe Li, Shen Gao, Xiangliang Zhang, Dongyan Zhao and Rui Yan. Capturing Relations between Scientific Papers: An Abstractive Model for Related Work Section Generation. To appear at the ACL-IJCNLP 2021​ main conference.

• Zhenwen Liang, Xiangliang Zhang.  Solving Math Word Problems with Teacher Supervision. Accepted by The 30th International Joint Conference on Artificial Intelligence (IJCAI 2021), Montreal-themed Virtual Reality, 21st -26th August, 2021​ (acceptance rate of 13.9%, 587/4204)

• Xiangliang Zhang, Qiang Yang, Somayah Albaradei, Xiaoting Lyu, Hind Alamro, Adil Salhi, Changsheng Ma, Manal Alshehri, Inji Ibrahim Jaber, Faroug Tifratene, Wei Wang, Takashi Gojobori, Carlos M. Duarte, Xin Gao. Rise and Fall of the Global Conversation and Shifting Sentiments During the COVID-19 Pandemic.  Nature’s Humanities and Social Sciences Communications.​ 2021. [PDF​]

• Zhenwen Liang, Jipeng Zhang, Jie Shao, Xiangliang Zhang. MWP-BERT: A Strong Baseline for Math Word Problems. CoRR abs/2107.13435 (2021)

• Basma Alharbi, Hind Alamro, Manal Alshehri, Zuhair Khayyat, Manal Kalkatawi, Inji Ibrahim Jaber, Xiangliang Zhang. ASAD: A Twitter-based Benchmark Arabic Sentiment Analysis Dataset.CoRR abs/2011.00578 (2020)