Scholar
Tianhao Wu
Google Scholar ID: df-THM0AAAAJ
University of California, Berkeley
Citations & Impact
Citations
1,990
H-index
13
i10-index
14
Publications
20
Co-authors
0
Publications
Resume (English only)
Academic Achievements
- Publications: 'Thinking LLMs: General Instruction Following with Thought Generation' (arXiv:2410.10630, 2024); 'EmbedLLM: Learning Compact Representations of Large Language Models' (arXiv:2410.02223, 2024); 'Meta-Rewarding Language Models: Self-Improving Alignment with LLM-as-a-Meta-Judge' (arXiv:2407.19594, 2024); 'Starling-7B: Improving LLM Helpfulness & Harmlessness with RLAIF' (Nov 2023).
Research Experience
- Research Intern at Meta in Summer 2024; Working on AI Society, a project that aims to form a more capable collective intelligence through modular connections of agents, potentially mitigating the computing demands limiting centralized AI systems today.
Education
- Degree: Ph.D.; University: UC Berkeley; Advisors: Jiantao Jiao, Kannan Ramchandran; Time: Ongoing; During undergrad, worked with Liwei Wang, majored in Mathematics.
Background
- Research Interests: Improving LLMs' instruction following and reasoning capabilities via (Self-Play) RL; Major: Mathematics; Brief Introduction: A 4th-year Ph.D. student at UC Berkeley, aiming to construct large-scale models capable of solving complex tasks requiring multi-step reasoning.
Miscellany
- No information provided regarding personal interests or hobbies.
Co-authors