Vrushabh27MIT
Phoenix — MolmoSpaces Benchmark Evaluation Artifacts
Complete evaluation artifacts for Phoenix, a fluid model for physical intelligence, evaluated on the MolmoSpaces benchmark suite using a Franka arm with joint-position action space. Evaluations conducted June–July 2026 with comprehensive per-episode videos and official scoring.
Downloads4K
Technical Profile
- Modalities
- rgbdepth
- Robot Embodiments
- Franka Panda
- Environment
- simulation
- Task Types
- pick_and_placemanipulationgraspingopeningclosing
- Data Format
- mp4 videos, HDF5, JSON manifest
- License
- MIT
Community Signals
Top 25% by downloads
Access
Need custom rgb data?
Claru builds purpose-built datasets for simulation applications with dense human annotations and quality assurance.
Request a Sample Pack