Jiakang Yuan

Ph.D. Candidate

About Me

I am current a fourth-year Ph.D. student (Sep. 2022 - Jun. 2027, expected) in the School of Information Science and Technology, Fudan University, supervised by Prof. Tao Chen. Before this, I obtained my Bachelor’s degree in Electronic Engineering also from Fudan University (Sep. 2018 - Jun. 2022). I’m currently working in the fields of agentic model, multi-agent system, and multi-modal large language models. My research pursues to develop AI systems that can understand, reason, and interact with the real world.

Download CV

Interests

Long-Horizon Agent Multi-Agent System Multimodal LLM Auto-Research

News

2026.08: Qwen3.8-Max, a 2.4T-parameter model with 95B active parameters, is released with strong improvements in coding, research, and long-horizon autonomous tasks.
2026.06: Agents-A1, an open-source 35B MoE model specialized on long-horizon tasks is out on Github and Huggingface, checkout the technical report.
2026.06: One papers (T^2VLA) is accepted by ECCV 2026 which is about test-time learning of VLA models.
2026.04: Three papers (MME-Reasoning, VisualScore, and, SAPO) are accepted by ICML 2026. Two of them are about multimodal reasoning, another is about visual quality assessment.
2026.04: Bi3D++ is accepted by IEEE T-PAMI 2026.
2026.04: Our survey about reward hacking is out on arxiv and github. Feel free to check it.
2026.04: Two papers (FlowSearch and Controllable Memory Usage) are accepted by ACL 2026. One is about DeepResearch, the other is about agent memory.
2026.03: Intern-S1-Pro technical report is released.
2026.02: InternAgent-1.5 technical report is released.
2025.12: SciEvalKit (An Open-source Evaluation Toolkit for Scientific General Intelligence) is released.
2025.10: Codes of InternAgent 1.0 is released on github.

Show more ↓

2025.07: SPOT is accepted by IEEE T-PAMI 2025.
2025.06: Two papers (Lumina Image 2.0 and Chimera) are accepted by ICCV 2025.
2025.05: Two papers (SurveyForge and Dolphin) are accepted by ACL 2025.
2025.05: We release InternAgent (NovelSeek), a unified closed-loop multi-agent framework for Automatic Scientific Research.
2025.02: One paper (CST-Stereo) is accepted by CVPR 2025.
2024.12: One paper (GeoX) is accepted by ICLR 2025
2024.12: One paper (AIOStereo) is accepted by AAAI 2025.
2024.10: I receive the National Scholarship.
2024.09: Two papers (AdaptiveDiffusion and 3DET-Mamba) are accepted by NeurIPS 2024.
2024.07: One paper (Reg-TTA3D) is accepted by ECCV 2024.
2024.01: One paper (ReSimAD) is accepted by ICLR 2024.
2023.09: One paper (AD-PT) is accepted by NeurIPS 2023.
2023.02: Two papers (Bi3D and Uni3D) are accepted by CVPR 2023.
2022.07: One paper (HelixFormer) is accepted by ACM’MM 2022.