Local goal support in policy functional graphs predicts goal failure in sparse GCRL with F1 0.925, and a taxonomy explains residual failures from competing attractors.
16 Proof.The first statement follows from Proposition 2: any trajectory reaching g from a non-goal start must enter g from some valid neighbor s∈N(g)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Functional Graphs for Predicting and Explaining Goal Failure in Sparse Goal-Conditioned RL
Local goal support in policy functional graphs predicts goal failure in sparse GCRL with F1 0.925, and a taxonomy explains residual failures from competing attractors.