Per-Context Reverse Curricula for Sparse-Reward Reinforcement Learning

Name: Per-Context Reverse Curricula for Sparse-Reward Reinforcement Learning
Uploaded: 2026-10-15T06:56:57.315922+00:00
Duration: 227 s
Description: Sparse-reward reinforcement learning often fails because training rollouts rarely reach later task stages. The SCOUT controller introduces an online, learner-agnostic reset mechanism that gives every context its own curriculum by deciding when to start training from easier intermediate states (scaffolds) and how quickly to remove assistance based on binary rollout success.

Sparse-reward reinforcement learning often fails because training rollouts rarely reach later task stages. The SCOUT controller introduces an online, learner-agnostic reset mechanism that gives every context its own curriculum by deciding when to start training from easier intermediate states (scaffolds) and how quickly to remove assistance based on binary rollout success.

Open the full episode