OmniToM evaluates Theory of Mind in large language models through explicit belief-structure modeling. Instead of scoring only endpoint answers, it asks models to extract actor-specific belief propositions from stories and label them along a seven-dimensional schema. The benchmark contains 895 stories and 22,343 labeled belief propositions, revealing an actor-specific information-tracking bottleneck in current LLMs.
@article{bawatneh2026omnitom,title={OmniToM: Benchmarking Theory of Mind in LLMs via Explicit Belief Modeling},author={Bawatneh, Adam and Sapkota, Sagar and Bedi, Amrit Singh and Karmaker, Santu and Shah, Mubarak},journal={arXiv preprint arXiv:2605.26322},year={2026},}
