ILQ clips model-imagined values of out-of-distribution actions at the maximum behavior value to limit optimism, reporting strong D4RL results, but its contraction proof is incorrect.
Diffusion policies creating a trust region for offline reinforcement learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Imagination-Limited Q-Learning for Offline Reinforcement Learning
ILQ clips model-imagined values of out-of-distribution actions at the maximum behavior value to limit optimism, reporting strong D4RL results, but its contraction proof is incorrect.