IAAIRLTT
Prioritize the Process, Not Just the Outcome: Rewarding Latent Thought Trajectories Improves Reasoning in Looped Language Models
Reward Latent Thought Trajectories distributes RL reward across the full latent reasoning trajectory of a looped model instead of only the final latent state, as a drop-in replacement for GRPO.
Inside the method
[ Shared stack ] × R → reward assigned across all latent steps
Simplified conceptual schematic. Consult the paper for the complete architecture.
- Recurrence family
- Full-stack recurrence
- Depth control
- Fixed loops of the base model
- KV / state strategy
- See paper
Reading note
On Ouro-1.4B/2.6B-Thinking it improves mean accuracy on MATH-500, AIME24/26 and BeyondAIME by 5.8% and 10.9% over GRPO, and transfers to non-math reasoning despite math-only training.
Sources checked 2026-10-28. This catalog does not imply independent reproduction.
Related works
Cite this work
@misc{rltt2026,
title = {Prioritize the Process, Not Just the Outcome: Rewarding Latent Thought Trajectories Improves Reasoning in Looped Language Models},
author = {Jonathan Williams and Esin Tureci},
year = {2026},
eprint = {2602.10520},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2602.10520}
}