Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Learning More Expressive General Policies for Classical Planning Domains

About

GNN-based approaches for learning general policies across planning domains are limited by the expressive power of $C_2$, namely; first-order logic with two variables and counting. This limitation can be overcame by transitioning to $k$-GNNs, for $k=3$, wherein object embeddings are substituted with triplet embeddings. Yet, while $3$-GNNs have the expressive power of $C_3$, unlike $1$- and $2$-GNNs that are confined to $C_2$, they require quartic time for message exchange and cubic space to store embeddings, rendering them infeasible in practice. In this work, we introduce a parameterized version R-GNN[$t$] (with parameter $t$) of Relational GNNs. Unlike GNNs, that are designed to perform computation on graphs, Relational GNNs are designed to do computation on relational structures. When $t=\infty$, R-GNN[$t$] approximates $3$-GNNs over graphs, but using only quadratic space for embeddings. For lower values of $t$, such as $t=1$ and $t=2$, R-GNN[$t$] achieves a weaker approximation by exchanging fewer messages, yet interestingly, often yield the expressivity required in several planning domains. Furthermore, the new R-GNN[$t$] architecture is the original R-GNN architecture with a suitable transformation applied to the inputs only. Experimental results illustrate the clear performance gains of R-GNN[$1$] over the plain R-GNNs, and also over Edge Transformers that also approximate $3$-GNNs.

Simon St\r{a}hlberg, Blai Bonet, Hector Geffner• 2024

Related benchmarks

TaskDatasetResultRank
Generalized PlanningIPC Ferry 2023 (test)
Coverage100
12
Generalized PlanningIPC Transport 2023 (test)
Coverage96
12
Generalized PlanningIPC Blocksworld 2023 (test)
Coverage98
12
Generalized PlanningIPC Satellite 2023 (test)
Coverage56
12
Generalized PlanningIPC Rovers 2023 (test)
Coverage32
12
Generalized PlanningIPC Floortile 2023 (test)
Coverage0.00e+0
12
Generalized PlanningIPC Childsnack 2023 (test)
Coverage40
11
PlanningIPC spanner 2023 (test)
Coverage97
8
PlanningIPC miconic 2023 (test)
Coverage90
8
PlanningIPC sokoban 2023 (test)
Coverage8
8
Showing 10 of 18 rows

Other info

Follow for update