Reasoning moves from visible tokens into latent space

LLoCoT matches chain-of-thought accuracy on code while cutting time to the first answer token about 36x, and a Nature Machine Intelligence perspective argues reasoning tokens are computational state, not narrative.

LLoCoT replaces chain-of-thought with latent loops · Reasoning tokens as computational state

Latent and recurrent reasoning methods move from research into inference stacks as a way to keep chain-of-thought accuracy while removing its token-latency tax, with faithfulness and interpretability of the hidden state becoming the next open question. Oct 9