Jiaheng Liu 刘佳恒

Assistant Professor · Ph.D. Supervisor

School of Intelligence Science and Technology
Nanjing University (Suzhou Campus)

NJU LINK Lab

Email: liujiaheng@nju.edu.cn

Jiaheng Liu

Biography

I am an Assistant Professor at Nanjing University, where I am a member of the Large-scale Intelligence and Knowledge (LINK) Lab.

My research spans foundation models, code intelligence, reinforcement learning, agents, and evaluation. I aim to build capable and reliable models that can reason, use tools, and complete complex tasks. My work connects open model development with post-training, realistic benchmarks, and applications in education and scientific research.

Previously, I was a Research Scientist at Alibaba (2023–2025), through the Alibaba Star program. I received my Ph.D. (2023) and B.Eng. (2019) in Software Engineering from Beihang University, advised during my Ph.D. by Prof. Ke Xu and Prof. Dong Xu. I also interned at SenseTime, Baidu, and Shanghai AI Laboratory.

I received an ACL 2024 Outstanding Paper Award. I contribute to open-source models and training frameworks, including MAP-Neo, OpenCoder, YuE, ROLL, and Flow-GRPO.

I serve as an Area Chair for ICLR, KDD, and ACL ARR and organize workshops at ICLR and AACL. See my professional services for details.

Research Interests

  • Foundation models: open pretraining, code and multimodal models, long-context modeling, and efficient inference.
  • Post-training and reinforcement learning: alignment, reasoning, reward modeling, and optimization for multimodal generation.
  • Agents and code intelligence: software engineering, tool use, deep research, and long-horizon task execution.
  • Evaluation: knowledge, factuality, safety, multimodal understanding, and realistic agent capabilities.

Explore my evaluation benchmarks by capability, or see my publications for recent work.

News

  • Three papers were accepted to NeurIPS 2026.
  • Twelve papers (7 Main and 5 Findings) were accepted to EMNLP 2026.
  • Nine papers were accepted to ICML 2026.
  • Twelve papers (5 Main and 7 Findings) were accepted to ACL 2026.
  • Thirteen papers were accepted to ICLR 2026.
  • Two papers were accepted to AAAI 2026.
Earlier news
  • Six papers were accepted to NeurIPS 2025.
  • Three papers were accepted to EMNLP 2025.
  • One paper was accepted to ACM MM 2025.
  • Thirteen papers (10 Main and 3 Findings) were accepted to ACL 2025.
  • Four papers were accepted to ICLR 2025.
  • Two papers were accepted to NAACL 2025.
  • Co-organizing the Open Science for Foundation Models Workshop at ICLR 2025.
  • Two papers were accepted to AAAI 2025.
  • Four papers were accepted to NeurIPS 2024.
  • One paper was accepted to EMNLP 2024.
  • Three papers were accepted to ACM MM 2024.
  • One paper was accepted to COLM 2024.
  • One paper was accepted to ECCV 2024.
  • Seven papers were accepted to ACL 2024.
  • One paper was accepted to ICML 2024.
  • One paper was accepted to CVPR 2024.
  • One paper was accepted to COLING 2024.
  • One paper was accepted to AAAI 2024.
  • One paper was accepted to Findings of EMNLP 2023.
  • One paper was accepted to ICCV 2023.
  • One paper was accepted to CIKM 2023.
  • One paper was accepted to Findings of ACL 2023.
  • One paper was accepted to CVPR 2023.
  • One paper was accepted to DASFAA 2023.
  • One paper was published in Multimedia Tools and Applications in 2023.
  • Two papers were accepted to ECCV 2022.
  • One paper was accepted to EMNLP 2022.
  • One paper was accepted to Findings of NAACL 2022.
  • One paper was accepted to AAAI 2022.
  • One paper was accepted to MICCAI 2022.
  • One paper was accepted to BIBM 2022.
  • One paper each was published in IEEE TIP, IEEE TMM, and IEEE TCSVT in 2022.
  • One paper was accepted to ICCV 2021.
  • Two papers were published in IEEE TCSVT in 2021.
  • One paper was accepted to AAAI 2020.
  • One paper was published in IEEE TIP in 2020.
  • Two papers were accepted to ICCV 2019.

Experience & Education

2025–present
Nanjing UniversityAssistant Professor, School of Intelligence Science and Technology
2023–2025
AlibabaResearch Scientist · Alibaba Star
2019–2023
Beihang UniversityPh.D. in Software Engineering
2015–2019
Beihang UniversityB.Eng. in Software Engineering