ResearchBenchmarksTheory of Mind Question Answering on MuMa-ToMFollow93.5AccuracyHuman31.51647.60863.779.792Apr 20, 2026Evaluation ResultsMethodMethodLinksAccuracyHumanEvaluator LLM=GPT-4oEvaluator LLM=GPT-4o2026.0493.5PDDL-MINDEvaluator LLM=GPT-4oEvaluator LLM=GPT-4o2026.0488.8AutoToMEvaluator LLM=GPT-4oEvaluator LLM=GPT-4o2026.0481.4LIMPEvaluator LLM=GPT-4oEvaluator LLM=GPT-4o2026.0476.6BIP-ALMEvaluator LLM=GPT-4oEvaluator LLM=GPT-4o2026.0433.9