Mahsa Bastankhah
@mahsabastankhah.bsky.social
PhD Student @Princeton, working on RL
In this paper, we provide a mechanistic explanation for how an RL agent can reach hard, faraway goals without any supervision. Please stop by our poster.
🔗Website: mahsa-bastankhah.github.io/demystifying...
🔗Website: mahsa-bastankhah.github.io/demystifying...
Demystifying Emergent Exploration in Goal-Conditioned RL
Understanding the mechanisms behind emergent exploration in unsupervised reinforcement learning
mahsa-bastankhah.github.io
December 5, 2025 at 5:35 PM
In this paper, we provide a mechanistic explanation for how an RL agent can reach hard, faraway goals without any supervision. Please stop by our poster.
🔗Website: mahsa-bastankhah.github.io/demystifying...
🔗Website: mahsa-bastankhah.github.io/demystifying...