About Me
I am currently an M.E. student in Computer Science and Technology at the School of Information, Renmin University of China, advised by Prof. Qin Jin and Asst. Prof. Wenxuan Wang. I received my B.E. in Computer Science and Technology from Renmin University of China in 2024.
I am also an intern student at Alibaba Cloud, supervised by Chenyang Lyu and Longyue Wang.
Research Interests:
- Multimodal Large Language Models — Building multimodal systems for understanding and reasoning across images, videos, and audio
- AI Alignment and Safety — Investigating how (M)LLMs and generative models align with human preferences and behave in terms of safety, fairness, and bias
- AI Applications for Sports — Developing AI models and systems that help fans better understand sports games and support officials in making accurate and consistent decisions
🔥 News
- 2026.06: Joined Alibaba Cloud as an intern student.
- 2026.04: POLYCHARTQA was accepted to ACL 2026 Main.
- 2026.04: A Survey of Large Models in Sports was accepted to ACL 2026 Findings.
- 2025.11: ChartEditor was accepted to AAAI 2026.
- 2025.07: ChartM3 was accepted to ACM MM 2025.
- 2024.09: TinyChart was accepted to EMNLP 2024.
📖 Education
-
Sep. 2024 - Present | M.E. in Computer Science and Technology
School of Information, Renmin University of China, Beijing, China
Advised by Prof. Qin Jin and Asst. Prof. Wenxuan Wang
-
Sep. 2020 - Jun. 2024 | B.E. in Computer Science and Technology
School of Information, Renmin University of China, Beijing, China
Advised by Prof. Qin Jin
💻 Research & Work Experience
-
Jun. 2026 - Present | Alibaba Cloud
Intern Student
Supervised by Chenyang Lyu and Longyue Wang
-
Jan. 2024 - Present | AIM3 Lab, Renmin University of China
Student Researcher, Beijing, China
Selected Projects
-
Multilingual Chart Understanding Project Lead Jan. 2025 - Jun. 2025
Built PolyChartQA, the first large-scale multilingual chart question-answering benchmark, through an automated data pipeline covering 10 languages and 16 chart types. Evaluation of more than 20 mainstream MLLMs revealed substantial gaps in low-resource languages; fine-tuning Qwen2.5-VL-7B achieved state-of-the-art performance. Accepted to ACL 2026 Main.
-
Awesome LLMs for Sports Project Lead Jul. 2025 - Nov. 2025
Conducted the first comprehensive survey of large language and multimodal models in sports, covering 19 tasks across 6 participant groups and synthesizing more than 240 papers. Released an open-source resource hub. Accepted to ACL 2026 Findings.
-
Automatic Sports Refereeing Project Lead Nov. 2025 - Present
Built RefereeBench, a professional-referee-annotated benchmark spanning 11 sports, 7 question types, 900+ videos, and 6,000+ question-answer pairs. The benchmark exposes a clear gap between frontier MLLMs and human experts in applying rules to realistic officiating scenarios.
📝 Publications
* denotes equal contribution. Selected publications and ongoing work.
Published
-
POLYCHARTQA: Benchmarking Large Vision-Language Models with Multilingual Chart Question Answering
Yichen Xu*, Liangyu Chen*, Liang Zhang, Jianzhe Ma, Wenxuan Wang, Qin Jin.
ACL 2026 Main (CCF A)
-
A Survey of Large Models in Sports
Yichen Xu*, Jianzhe Ma*, Chuhan Wang, Zhonghao Cao, Liangyu Chen, Wenxuan Wang, Qin Jin.
ACL 2026 Findings (CCF A)
-
ChartEditor: A Reinforcement Learning Framework for Robust Chart Editing
Liangyu Chen*, Yichen Xu*, Jianzhe Ma, Yuqi Liu, Donglu Yang, Liang Zhang, Wenxuan Wang, Qin Jin.
AAAI 2026 (CCF A)
-
ChartM3: Benchmarking Chart Editing with Multimodal Instructions
Donglu Yang, Liang Zhang, Zihao Yue, Liangyu Chen, Yichen Xu, Wenxuan Wang, Qin Jin.
ACM MM 2025 (CCF A)
-
TinyChart: Efficient Chart Understanding with Visual Token Merging and Program-of-Thoughts Learning
Liang Zhang, Anwen Hu, Haiyang Xu, Ming Yan, Yichen Xu, Qin Jin, Ji Zhang, Fei Huang.
EMNLP 2024 (CCF B)
Preprints & Under Review
-
RefereeBench: Are Video MLLMs Ready to be Multi-Sport Referees?
Yichen Xu, Yuanhang Liu, Chuhan Wang, Zihan Zhao, Jinghan Luo, Jianzhe Ma, Wenxuan Wang, Qin Jin.
Under review by ACM MM 2026
-
EgoEsportsQA: An Egocentric Video Benchmark for Perception and Reasoning in Esports
Jianzhe Ma, Zhonghao Cao, Shangkui Chen, Yichen Xu, Wenxuan Wang, Qin Jin.
Under review by ACM MM 2026
-
SPA-Bench: A Novel Video Benchmark for Proactive Physical Hazard Prediction in MLLMs
Jiawei Qiu*, Yichen Xu*, Jianzhe Ma, Mingyang Yu, Wenbin Zhu, Yang Han, Pinzheng Lv, Wenxuan Wang.
Under review by ACM MM 2026
-
THVL-Bench: A Multimodal Benchmark for Temporal Localization of Harmful Content in Long Videos
Wenjie Jiang*, Yichen Xu*, Chenrui Zhu, Minghao Yan, Yiming Chang, Jiawei Qiu, Chengwei Qin, Wenxuan Wang.
Under review by NeurIPS 2026
-
CoachBench: Benchmarking Sports Coaching Knowledge of Large Language Models
Yuanchun Wang, Yuzhe Xin, Jianzhe Ma, Ao Guan, Mingqiu Dai, Yichen Xu, Xi Qin, Jiayu Wang, Wenxuan Wang, Jing Zhang.
Under review by NeurIPS 2026
-
Manners Maketh Man: Evaluating Text-to-Image Models on Everyday Safety and Social Etiquette
Yichen Xu, Tianhuiyu Zou, Wenxuan Wang.
Under review by ARR May 2026
-
FineFencing: A Fine-Grained Sports Video Benchmark for Fencing Understanding
Yichen Xu, Jinghan Luo, Shunran Wang, Pinzheng Lv, Zilu Qiu, Pinqi Xie, Jianzhe Ma, Wenxuan Wang.
Under review by ARR May 2026
-
Beyond Captioning: Evaluating Sports Commentary Generation through Human Commentary Behavior Distributions
Yichen Xu, Yan Shi, Junnan Hu, Shuangqi Wu, Rongtao Chen, Mohan Zhang, Shangkui Chen, Xi Qin, Wenxuan Wang, Jianzhe Ma.
Under review by ARR May 2026
-
IceArtBench: Benchmarking Expert-Level Aesthetic Appreciation of Figure Skating in MLLMs
Jianzhe Ma, Yichen Xu, Wenxuan Wang.
Under review by ARR May 2026
🛠 Skills
- Languages: Chinese, English (IELTS 8.0)
- Programming: Python, Matlab, C/C++, Java
- Technologies: PyTorch, TensorFlow, OpenCV, Linux