Yongji Fu

Yongji Fu 符永骥

I am Yongji Fu (符永骥), an MSc student in Robotics Engineering at the University of Bristol (2025.09 – 2026.10), advised by Nathan F. Lepora and Guanqun Cao. I was also a Research Assistant at the Institute of Engineering and Applied Technology, Fudan University (2025.02 – 2026.03), working on realistic expression learning for humanoid face robots and on post-training and evaluation of computer-use (GUI) agents. Before Bristol I received my BSc in Information Management and Information Systems from Chongqing University of Posts and Telecommunications.

Goal   To build robotic and agentic systems that continuously learn and iteratively self-improve through interaction with the physical world.

Research Interests   large-scale reinforcement learning ·world model for robot learning · continuous self-evolving agent · general-purpose loco-manipulation

ROSE: A Shared Cross-Embodiment Facial-Motion Latent Space for Humanoid Robot Expression

Yongji Fu, Man Cao, Yi Zhou, Fei Lin, Mark F. Hansen, Guanqun Cao

ICRA 2027, Submitted, 2026

TouchSteer: Grounding Natural Language in Tactile Perception via Steering Vectors

Guanqun Cao, Yongji Fu, Yi Zhou, Gaojie Jin, Zhenyu Lu, Shan Luo

IEEE Transactions on Robot Learning, Under review, 2026

Constructing Dynamic S-boxes Based on Chaos and Irreducible Polynomials for Image Encryption

Constructing Dynamic S-boxes Based on Chaos and Irreducible Polynomials for Image Encryption

Chenhong Luo, Yong Wang, Yongji Fu, et al.

Nonlinear Dynamics(Springer,JCR Q1,IF 6.0), 2024

Computer-Use GUI Agent: Post-Training and Self-Evolving Evaluation

Computer-Use GUI Agent: Post-Training and Self-Evolving Evaluation

Post-training and evaluation for a computer-use agent that runs multi-step workflows on real web and desktop interfaces. On Qwen2.5-14B-Instruct, an SFT → offline step-RFT → online GiGPO pipeline lifts success on a 200-task expert-solution benchmark from 31.3% to 49.2%, and multi-teacher on-policy distillation (MOPD) then merges GiGPO specialists for macOS, Linux, Windows, and Web into a single cross-platform agent. A trace-based LLM-as-Judge evaluator agrees with human labels 94.3% of the time (83.2% on complex tasks) and drives a self-evolving benchmark.

  • GUI Agent
  • Qwen2.5
  • SFT
  • GRPO
  • GiGPO
  • DAPO
  • MOPD
  • LLM-as-Judge

Continually Learning Interactive Robot

An embodied agent that keeps expanding its behavior repertoire through ongoing human–robot interaction — new skills, new object concepts, and new language grounding are acquired online rather than baked in at training time.

  • HRI
  • Continual Learning
  • Multimodal
  • Agent
Thinking-with-Image: RL for Tool-Augmented Visual Reasoning

Thinking-with-Image: RL for Tool-Augmented Visual Reasoning

Reproduced and extended the VTool-R1 / ReFocus line of work so a VLM actively rewrites the chart it is reading — masking, boxing, and highlighting the relevant rows and columns mid-reasoning — turning thinking-in-text into thinking-in-image. The tool-using policy is trained end-to-end with RL (veRL + GRPO + DAPO). On ChartQA-style QA with Qwen2.5-VL-3B, answer accuracy rose from 62% to 96%.

  • RL
  • veRL
  • DAPO
  • GRPO
  • Qwen2.5-VL
  • vLLM