Scholar
Shibo Hao
Google Scholar ID: xwbHbUQAAAAJ
Ph.D. student, UC San Diego
Citations & Impact
Citations
1,586
H-index
11
i10-index
11
Publications
17
Co-authors
9
list available
Publications
Resume (English only)
Academic Achievements
- Publications: 'Reasoning by Superposition: A Theoretical Perspective on Chain of Continuous Thought' (NeurIPS 2025), 'Offline Reinforcement Learning for LLM Multi-Step Reasoning' (ACL 2025), 'Revisiting Reinforcement Learning for LLM Reasoning from A Cross-Domain Perspective' (NeurIPS 2025), 'Training Large Language Models to Reason in a Continuous Latent Space' (COLM 2025), etc. Awards: ToolkenGPT received the best paper award at NeurIPS 2023.
Research Experience
- Research scientist intern at Meta FIAR lab, mentored by Yuandong Tian and Jason Weston. Involved in multiple research projects such as Guru, OREO, FoR, Coconut, etc.
Education
- Ph.D. student at UC San Diego, advised by Zhiting Hu; B.S. in Computer Science from Peking University.
Background
- Research interests: machine reasoning. Work includes training large language models to reason with reinforcement learning, exploring reasoning in latent space, building a system-2 reasoning framework using world-model planning, and augmenting LLMs with external tools.
Miscellany
- Personal interests not mentioned.
