AGMCTS augments MCTS with action-score gradients for particle beliefs, a Multiple Importance Sampling tree for reuse, and Area Formula gradients for smooth models, outperforming prior sample-based solvers on continuous benchmarks.
For AGMCTS, we also optimized over the Adam learning rate ηAdam, and the action update distance threshold T min da only in the Light-Dark domains
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Action-Gradient Monte Carlo Tree Search for Non-Parametric Continuous (PO)MDPs
AGMCTS augments MCTS with action-score gradients for particle beliefs, a Multiple Importance Sampling tree for reuse, and Area Formula gradients for smooth models, outperforming prior sample-based solvers on continuous benchmarks.