Transformers develop four algorithmic phases of in-context learning on Markov chains via two distinct multi-layer subcircuit mechanisms, with phase boundaries set by data diversity K.
This allows us to replace y(1) n =x (0) n + Att(1) x(0) ≤n − →y (1) n =x (0) n ⊕x (0) n−1,(XII7) where⊕denotes concatenation into orthogonal subspaces
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Distinct mechanisms underlying in-context learning in transformers
Transformers develop four algorithmic phases of in-context learning on Markov chains via two distinct multi-layer subcircuit mechanisms, with phase boundaries set by data diversity K.