P2N: Reusing Deep Representations for Greater Effective Depth
Feeds the previous token’s deep hidden state into earlier layers for the next token. A shared middle block and parallel Jacobi updates train this recurrence without serializing all training tokens.
First public: · Added to this list: November 2, 2026
Explore this approach
Recurrent reasoning — Repeat parts of the model or keep intermediate information in memory while working through a problem. Some methods also learn how many passes to run.