Regularisation on Mohamad H. Daneshhttps://modanesh.github.io/tags/regularisation/Recent content in Regularisation on Mohamad H. DaneshHugo -- 0.147.2enWed, 07 Oct 2026 00:00:00 +0000Regularised RL Improves Sim-to-Real Transferhttps://modanesh.github.io/papers/smm/Wed, 07 Oct 2026 00:00:00 +0000https://modanesh.github.io/papers/smm/SMM-AC extends the soft mellowmax value function to continuous control. On real Franka and ANYmal-D robots it transfers better from simulation and keeps more of its return under deployment-time perturbations than PPO, FastTD3, and SAC.