Scholar
Hao Peng
Google Scholar ID: 6Y37nm0AAAAJ
University of Illinois Urbana-Champaign
Citations & Impact
Citations
6,203
H-index
36
i10-index
58
Publications
20
Co-authors
0
Publications
Resume
Academic Achievements
- 2025: 'From f(x) and g(x) to f(g(x)): LLMs Learn New Skills in RL by Composing Old Ones'
- 2025: Co-authored 'Probing the Critical Point (CritPt) of AI Reasoning: a Frontier Physics Research Benchmark' (58 authors total)
- 2025: 'Executable Counterfactuals: Improving LLMs’ Causal Reasoning Through Code'
- 2025: 'Context Length Alone Hurts LLM Performance Despite Perfect Retrieval' (EMNLP Findings)
- 2025: 'The Best Instruction-Tuning Data are Those That Fit' (NeurIPS)
- 2025: 'The Unreasonable Effectiveness of Entropy Minimization in LLM Reasoning' (NeurIPS)
- 2025: 'Reinforcement Learning Finetunes Small Subnetworks in Large Language Models' (NeurIPS)
Background
- Assistant Professor at the Department of Computer Science, University of Illinois at Urbana-Champaign (UIUC)
- Current research focuses on large language models (LLMs)
- Works on solving complex reasoning problems in a generalizable way, emphasizing learning from experience (e.g., reinforcement learning) and insights from human cognition
- Interested in causal understanding and reasoning about the world
- Committed to positively impacting society through AI
- Aims to advance the frontier of human knowledge and contribute to scientific discovery as the ultimate demonstration of true generalization beyond training data