Adam: An Embodied Causal Agent in Open-World Environments

Shu Yu1, 2, Chaochao Lu2,†
1Fudan University 2Shanghai Artificial Intelligence Laboratory
Corresponding author.

The causal graph learned in lifelong learning.
Adam successfully unlocks all 41 actions we implement and discovers accurate causal relationships.

Abstract

mineDiamondOre

gatherWoodLog

gatherStone

smeltRawIron


Automatic exploration


Components




Environmental awareness

Causal graph learning

Task executing

Performance

Interpretability



Efficiency


Robustness


Conclusion


BibTeX

@article{yu2024adam,
  title   = {ADAM: An Embodied Causal Agent in Open-World Environments},
  author  = {Shu Yu and Chaochao Lu},
  year    = {2024},
  journal = {arXiv preprint arXiv:2410.22194}
}