Kitchen

The 'Kitchen' dataset is a benchmark that contains high-dimensional state representations used to evaluate the performance of algorithms in handling data corruption in offline reinforcement learning scenarios.

Papers using Kitchen (6)

ADG: Ambient Diffusion-Guided Dataset Recovery for Corruption-Robust Offline Reinforcement Learning2025Offline Reinforcement Learning with Discrete Diffusion Skills2025Preferred-Action-Optimized Diffusion Policies for Offline Reinforcement Learning2024PlanDQ: Hierarchical Plan Orchestration via D-Conductor and Q-Performer2024Tackling Data Corruption in Offline Reinforcement Learning via Sequence Modeling2024DIAR: Diffusion-model-guided Implicit Q-learning with Adaptive Revaluation2024