Yingru Li
Yingru Li
Home
Posts
Research
Contact
Resume
RL-Seminar
Light
Dark
Automatic
Language Models
Information Bandwidth in Reinforcement Learning
A mathematically rigorous information-theoretic analysis of learning efficiency in RL algorithms, explaining why LoRA works for policy gradient fine-tuning.
Yingru LI
Oct 1, 2025
11 min read
Research
,
Theory
Cite
×