Preprint
Jul 2026
Reasoning Fine-Tuning Induces Persistent Latent Policy States
This work modeling Chain-of-Thought reasoning as a switching dynamical system (SDS) suggests that reasoning fine-tuning globally reorganizes latent dynamics, offering a new lens for mechanistic analysis and process-level control of reasoning models.
Abir Harrasse, Michael Lan, Hunar Batra et al.
· 0 citations