TY  - RPRT
TI  - An Information-Theoretic Approach to Understanding Transformers' In-Context Learning of Variable-Order Markov Chains
AU  - Ruida Zhou
AU  - Chao Tian
AU  - Suhas Diggavi
PY  - 2026
UR  - https://arxiv.org/abs/2410.05493
ID  - 2410.05493
ER  -