About Me

I am a fourth-year PhD candidate in the Department of Computer Science and Engineering at The Chinese University of Hong Kong (CUHK), supervised by Prof. James Cheng. I am currently a Top Talent Intern at MiniMax, working with the Foundation Model Team’s Reinforcement Learning Group.

My research focuses on improving multi-turn coding agents through reinforcement learning (RL) and on-policy distillation (OPD).

Previously, I studied graph neural networks and attention mechanisms under the supervision of Dr. Yifei Wang at MIT CSAIL. I have also collaborated closely with Dr. Xinyi Wu at MIT IDSS and Dr. Kevin Qinghong Lin at Oxford Torr Visual Group.

Prior to coming to CUHK, I was an undergraduate student at Harbin Institute of Technology, where I worked as a research intern at SCIR, supervised by Prof. Libo Qin.

I welcome research collaborations in these areas. Feel free to contact me at wjqkoko@gmail.com.

I expect to graduate in Fall 2027 and am seeking full-time opportunities in LLM post-training and quantitative research starting in late 2027. Please feel free to reach out!

🔥 News

• 2026.09: 🎉 Our VideoASMR-Bench dataset has now surpassed 6k downloads.
• 2026.07: 🎉 Our TCOD is accepted by COLM 2026.
• 2026.05: 🎉 Our VideoASMR-Bench was accepted by CVPR 2026 VGBE Workshop. The dataset is publicly available.
• 2026.01: 🎉 Our Prost-LLM is accepted by ICASSP 2026.
• 2025.09: 🎉 Our three papers are accepted by NeurIPS 2025.
• 2025.08: 🎉 Our MLMT is accepted by 2025 IEEE ASRU.

🎓 Education

2023.09–Present, Ph.D. in Computer Science and Engineering @ The Chinese University of Hong Kong (expected Fall 2027).

2019.09–2023.09, B.Eng. in Artificial Intelligence @ Harbin Institute of Technology, Honors College.

💻 Internships

2026.06 — Present

MiniMax

Top Talent Intern Program · Beijing

Foundation Model Team · Reinforcement Learning Group

Working on MOPD and RL scaling, contributing to releases of the M-series LLMs.

2025.10 — 2026.04

Tongyi Lab, Alibaba Group

Research Intern · Hangzhou

Research on LLM reasoning, reinforcement learning, and on-policy distillation.

2022.05 — 2022.08

Microsoft

Engineer Intern · Shanghai

Azure Spring App.

📝 Publications

† indicates equal contribution.

First-author

	TCOD: Exploring Temporal Curriculum in On-Policy Distillation for Multi-turn Autonomous Agents Jiaqi Wang, Wenhao Zhang, Weijie Shi, Yaliang Li, James Cheng. COLM 2026 [paper] [code] [huggingface paper]
	VideoASMR-Bench: Can AI-Generated ASMR Videos Fool VLMs and Humans? Jiaqi Wang†, Weijia Wu†, Yi Zhan, Rui Zhao, Ming Hu, James Cheng, Wei Liu, Philip Torr, Kevin Qinghong Lin. CVPR 2026 Workshop VGBE [paper] [homepage] [code] [dataset] [huggingface paper]
	Think or Not? Selective Reasoning via Reinforcement Learning for Vision-Language Models Jiaqi Wang†, Kevin QH. Lin†, James Cheng, Mike Z. Shou. NeurIPS 2025, ICML 2025 EXAIT Workshop [paper] [code] [huggingface]
	A Signed Graph Approach to Understanding and Mitigating Oversmoothing in GNNs Jiaqi Wang†, Xinyi Wu†, James Cheng, Yifei Wang. NeurIPS 2025 [paper] [code]
	DivIL: Unveiling and Addressing Over-Invariance for Out-of-Distribution Generalization Jiaqi Wang†, Yuhang Zhou†, Zhixiong Zhang†, Qiguang Chen, Yongqiang Chen, James Cheng. TMLR 2025 [paper] [code]

Non-first-author

	PROST-LLM: Progressively Enhancing the Speech-to-Speech Translation Capability in LLMs Jing Xu, Jiaqi Wang, Daxin Tan, Xiao Chen. ICASSP 2026 [paper]
	Visual Thoughts: A Unified Perspective of Understanding Multimodal Chain-of-Thought Zihui Cheng, Qiguang Chen, Xiao Xu, Jiaqi Wang, Weiyun Wang, Hao Fei, Yidong Wang, Alex Jinpeng Wang, Zhi Chen, Wanxiang Che, Libo Qin. NeurIPS 2025 [paper]
	Enhancing Multilingual Speech Generation and Recognition Abilities in LLMs with Constructed Code-switched Data Jing Xu, Daxin Tan, Jiaqi Wang, Xiao Chen. IEEE ASRU 2025 [paper]
	PIGDreamer: Privileged Information Guided World Models for Safe Partially Observable Reinforcement Learning Dongchi Huang, Jiaqi Wang, Yang Li, Chunhe Xia, Tianle Zhang, Kaige Zhang. ICML 2025 [paper] [code]
	Unlocking the capabilities of thought: A reasoning boundary framework to quantify and optimize chain-of-thought Qiguang Chen, Libo Qin, Jiaqi Wang, Jingxuan Zhou, Wanxiang Che. NeurIPS 2024, Oral [paper] [code]
	Enhancing evolving domain generalization through dynamic latent representations Binghui Xie, Yongqiang Chen, Jiaqi Wang, Kaiwen Zhou, Bo Han, Wei Meng, James Cheng. AAAI 2024, Oral [paper]

🏆 Honors

• 2025.10, Neurips Scholar Award.
• 2021.10, National Scholarship.