logo
today local_bar
In-Context Learning; Induction Head; Mechanistic Interpretability; Task Vector;

1 paper across 1 session

Poster Session 2

1 paper
Wednesday, December 3, 2025 · 4:30 PM → 7:30 PM
Exhibit Hall C,D,E
Unifying Attention Heads and Task Vectors via Hidden State Geometry in In-Context Learning
#1008 · Haolin Yang, Hakaze Cho, Yiqiao Zhong, Naoya Inoue

We unify attention heads and task vectors in the mechanistic analysis of in-context learning.