World Wires · story 104070 · corroborated · 1 source(s)

GSBR: Geometric Support Behavior Regularization for Online RL

Geometric support behavior regularization improves sample efficiency and stability in online reinforcement learning without overly restricting policy impro

Open in the desk

Coverage

• GSBR: Geometric Support Behavior Regularization for Online RL arXiv: new preprints (security, AI, physics) · tier C · clear net · 2026-11-03 18:08 UTC

What this site indexes