long-horizon tasks
Complex tasks that require long-term planning and consideration of multiple sequential decisions to achieve a goal, typical in reinforcement learning and robotics.
- Flattening Hierarchies with Policy Bootstrapping
- HumanoidGen: Data Generation for Bimanual Dexterous Manipulation via LLM Reasoning
- Matryoshka Pilot: Learning to Drive Black-Box LLMs with LLMs
- Mind2Web 2: Evaluating Agentic Search with Agent-as-a-Judge
- Option-aware Temporally Abstracted Value for Offline Goal-Conditioned Reinforcement Learning
- Physics-informed Value Learner for Offline Goal-Conditioned Reinforcement Learning
- Reinforcement Learning with Action Chunking
- SutureBot: A Precision Framework & Benchmark For Autonomous End-to-End Suturing