VirtualHome

The 'VirtualHome' dataset is a benchmark that contains simulated environments for evaluating reinforcement learning agents on complex, goal-oriented tasks requiring multi-step interactions.

Papers using VirtualHome (3)

SPA-RL: Reinforcing LLM Agents via Stepwise Progress Attribution2025LLM-Based Offline Learning for Embodied Agents via Consistency-Guided Reward Ensemble2024A Transformer-Based Model for the Prediction of Human Gaze Behavior on Videos2024