Abstract simulators can be grounded to real tasks by making their dynamics history-dependent and correcting them with real data, enabling RL policy transfer.
Dd-ppo: Learning near-perfect pointgoal navigators from 2.5 billion frames,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Abstract Sim2Real through Approximate Information States
Abstract simulators can be grounded to real tasks by making their dynamics history-dependent and correcting them with real data, enabling RL policy transfer.