NeoRL

Dataset Card for NeoRL‑2: Near Real‑World Benchmarks for Offline Reinforcement Learning Dataset Summary NeoRL-2 is a collection of seven near–real-world offline-RL datasets plus their evaluation simulators. This repo we provide the offline-RL dataset, while the simulators are in https://github.com/polixir/NeoRL2. Each

Papers using NeoRL (2)

Long-Horizon Model-Based Offline Reinforcement Learning Without Conservatism2025VIPO: Value Function Inconsistency Penalized Offline Reinforcement Learning2025