Offline Goal-Conditioned Reinforcement Learning with Projective Quasimetric Planning
This work addresses long-horizon goal-reaching problems in offline reinforcement learning, offering a solution for tasks like navigation, though it appears incremental as it builds on existing geometric and compositional approaches.
The paper tackles the challenge of scaling offline goal-conditioned reinforcement learning to long-horizon tasks by addressing compounding value-estimation errors, introducing Projective Quasimetric Planning (ProQ) to learn an asymmetric distance for generating sub-goals and guiding control, resulting in robust performance on diverse navigation benchmarks.
Offline Goal-Conditioned Reinforcement Learning seeks to train agents to reach specified goals from previously collected trajectories. Scaling that promises to long-horizon tasks remains challenging, notably due to compounding value-estimation errors. Principled geometric offers a potential solution to address these issues. Following this insight, we introduce Projective Quasimetric Planning (ProQ), a compositional framework that learns an asymmetric distance and then repurposes it, firstly as a repulsive energy forcing a sparse set of keypoints to uniformly spread over the learned latent space, and secondly as a structured directional cost guiding towards proximal sub-goals. In particular, ProQ couples this geometry with a Lagrangian out-of-distribution detector to ensure the learned keypoints stay within reachable areas. By unifying metric learning, keypoint coverage, and goal-conditioned control, our approach produces meaningful sub-goals and robustly drives long-horizon goal-reaching on diverse a navigation benchmarks.