2 problems
- 0 votes0 replies1 view
The separation conjecture
– separation conjecture. . Under this widely believed conjecture, chain-of-thought computation would extend the expres…
- 0 votes0 replies0 views
The robust predicate–attention conjecture for LEGO length generalization
Let be the length of a LEGO task, whose input consists of predicate and answer clauses, and let attention patterns describe which clauses a transformer uses when predicting the…