AI.info

Shawn Im

Explore Shawn Im on AI.info.

• How Do Transformers Learn to Associate Tokens: Gradient Leading Terms Bring Mechanistic Interpretability