Academic Blog of Hakaze Cho. He is a Ph.D. student focusing on Analyzing Deep Learning.

News

• My paper Revisiting In-context Learning Inference Circuit in Large Language Models has been accepted to ICLR 2025.
• My paper Token-based Decision Criteria Are Suboptimal in In-context Learning has been accepted to NAACL 2025-main.

Blogs

Articles: Research and Technique

Articles: Other

Archive: Old Articles from Previous Blog

Recommended Articles

Light-hearted Mechanistic Interpretability ABC

CV / Publications / Google Scholar

More Detailed Self-introduction

万物尽在你眼前

Illustrator: @ぽち @ふか @あめ飴