Do Agents Know When They Succeed? Calibrating Agent Confidence from Internal Representations
This paper investigates whether model's internal representations provide stronger signals of eventual task success in multi-turn agentic setups, and introduces two complementary methods: Latent Trajectory Dynamics (LTD) and the Action Representation Probe (ARP), which predicts success from representations formed at act...
Priyanka Mary Mammen, Emil Joswin, Srujananjali Medicherla
· 1 citation