MuJoCo

MuJoCo is a benchmark that contains a set of reinforcement learning tasks used to evaluate algorithms in the context of reinforcement learning from human feedback.

Papers using MuJoCo (3)

Efficient Federated RLHF via Zeroth-Order Policy Optimization2026dataRLsec: Safety, Security, and Reliability With Robust Offline Reinforcement Learning for DPAs2026Improved Communication Efficiency in Federated Natural Policy Gradient via ADMM-based Gradient Updates2023 · 10 cites