TY  - RPRT
TI  - Agentic In-Context Learning: What Algorithms Can Transformers Learn in Closed-Loop Interaction?
AU  - Shiguang Wu
AU  - Yaqing Wang
AU  - Quanming Yao
PY  - 2026
UR  - https://arxiv.org/abs/2610.23051
ID  - 2610.23051
ER  -