robotics
The Demonstration Includes Mistakes; the Progress Meter Does Not
PRIME links similar states across trajectories and measures distance to success instead of assuming every recorded action advances the task.
Summary
PRIME links similar states across trajectories and measures distance to success instead of assuming every recorded action advances the task.
A robot demonstration can contain pauses, corrections and wrong turns, yet many progress estimators treat the whole trajectory as monotonic success. PRIME builds a weighted graph across observed states, connects visually and spatially similar moments, and uses shortest paths to estimate distance to the goal. That lets the reward signal distinguish advancement, stagnation and setbacks. Across simulated and real precision-manipulation tasks, the authors report stronger robustness to suboptimal demonstrations and as much as a fivefold policy-success improvement over supervised fine-tuning baselines on charger insertion. The magnitude is task-specific, but the principle is broader: learning from a demonstration should not require pretending the demonstrator was perfect.
Why it matters
PRIME links similar states across trajectories and measures distance to success instead of assuming every recorded action advances the task.
Limits and context
- The magnitude is task-specific, but the principle is broader: learning from a demonstration should not require pretending the demonstrator was perfect.
Key claims
PRIME links similar states across trajectories and measures distance to success instead of assuming every recorded action advances the task.
Qualification: The magnitude is task-specific, but the principle is broader: learning from a demonstration should not require pretending the demonstrator was perfect.
Evidence: source-2026-10-19-010
Sources
- arXiv preprint 2610.23769, original v1arXiv · primary research
Corrections
No corrections have been recorded for this story.