Reproduce “Tiresias: Towards Language-Conditioned World Models” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Fengyi Wu, Yifei Dong, Yetong Sha, Lingdong Kong, Qiyu Hu, Bohan Xiong, Qi Dai, Alexander G. Hauptmann and 1 more
Tiresias plans action sequences from an initial view and language instructions across navigation, robotics, 3D gaming, and driving. It samples 128 trajectories without scene input, filters them by motion and language, then imagines visual outcomes for eight shortlisted candidates to select a plan. On val seen, it scores 63% versus 54% for LCVN-WM, but real-world deployment remains unvalidated.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.