Flaw Detected in Commutative Factor Detection for Lifted Inference

A new arXiv preprint (2605.26908) reveals that the state-of-the-art algorithm for detecting commutative factors in factor graphs relies on a theorem that provides only a necessary condition, not a sufficient one as previously assumed. Commutative factors are central to lifted probabilistic inference, enabling tractable reasoning by exploiting object indistinguishability. The paper revisits the theoretical foundations and demonstrates that the current algorithm may produce false positives, potentially undermining inference correctness. The authors provide corrected necessary and sufficient conditions for commutative factor detection. No specific authors or institutions are named in the abstract.