Presents efficient holistic lookahead encoding and abstracted IW(1) that enable relational GNN policies to achieve new SOTA results surpassing prior work and LAMA on hyperscaling IPC 2023 benchmarks.
Learning general optimal policies with graph neural networks: Expressive power, transparency, and limits
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Efficient Lookahead Encoding and Abstracted Width for Learning General Policies in Classical Planning
Presents efficient holistic lookahead encoding and abstracted IW(1) that enable relational GNN policies to achieve new SOTA results surpassing prior work and LAMA on hyperscaling IPC 2023 benchmarks.