An Evolving Benchmark for Functional Theory of Mind in Embodied Agents.
EnactToM evaluates whether embodied agents can use beliefs about other agents to act correctly, not just answer direct belief questions. Agents operate in shared 3D household scenes with private observations, constrained communication, and formally verified goals.
conda create -n enacttom python=3.10 cmake=3.14.0 -y
conda activate enacttom
python -m pip install -r requirements.txt
python -m pip install -e .
./enacttom/run.sh --helpHabitat-backed generation, replay, and benchmarking require the simulator and assets described in the installation guide.
If our work was useful for you, please cite it:
@article{enacttom2026,
title={EnactToM: An Evolving Benchmark for Functional Theory of Mind in Embodied Agents},
author={Gurusha Juneja and Dylan Lu and Saaket Agashe and Parth Diwane and Edward Gunn and Jayanth Srinivasa and Gaowen Liu and William Yang Wang and Yali Du and Xin Eric Wang},
year={2026},
url={https://enact-tom.github.io}
}EnactToM is licensed under the terms of the license found in LICENSE.