SCORE constrains sim RL to the support of a real-data policy via flow steering, raising average success on eight dexterous tasks from 37.8% to 89.9%.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.RO 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
A contact-centric framework extracts contact event sequences from one demonstration to serve as structured reward for RL, yielding more stable sim-to-real transfer than unconstrained baselines in manipulation tasks.
citing papers explorer
-
Support-Constrained RL Enables Real-World Policy Improvement without Real-World Experience
SCORE constrains sim RL to the support of a real-data policy via flow steering, raising average success on eight dexterous tasks from 37.8% to 89.9%.
-
ConCent: Contact-Centric Real-to-Sim-to-Real Learning from One Demonstration
A contact-centric framework extracts contact event sequences from one demonstration to serve as structured reward for RL, yielding more stable sim-to-real transfer than unconstrained baselines in manipulation tasks.