Reward-DAgger: Robot-Gated Interactive Imitation Learning
Overview
Reward-DAgger uses progress-based reward models to trigger human intervention in imitation learning. The framework aims to improve robot autonomy through interactive learning.
Generated from attributed reports · 1 hours agoUpdated
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
No current evidence-backed claims. Missing information remains not reported.
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsReward-DAgger: Robot-Gated Interactive Imitation Learning with General-Purpose Progress-Based Reward Models
This paper presents Reward-DAgger, a robot-gated interactive imitation learning framework that uses dense progress signals from a general-purpose reward model to determine when human intervention is needed. The approach is agnostic to policy architecture, requires no access to policy internals, and achieves better failure-detection accuracy-latency tradeoffs than existing baselines across simulated and real-world tasks.
Event attention history
There is not enough continuous observation data to show a trend.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).