← Leaderboard

Seer

Policy
Shanghai AI Lab · Dec 19, 2024

End-to-end predictive inverse dynamics policy from Shanghai AI Lab with PKU and CUHK that forecasts future visual states and infers actions from them, pretrained on DROID. It reports CALVIN results for Seer and Seer-Large.

Type
Policy
Parameters
—
Backbone
Predictive inverse dynamics transformer (Seer-Large: 315M trainable + 251M frozen encoders)
Weights
Open
License
Apache-2.0
Commercial use
Allowed
Released
Dec 19, 2024
Robots
Franka, Franka Panda (sim)
SO-101
–
LeRobot
–
Inputs
–
VRAM
–

Benchmark results

1 of 13 benchmarks
CALVIN ABC→DNear saturationSimulation
#12 of 244.28 / 5, self-reported by the authors, not a leaderboard entry
Not reported on 12 benchmarks
Wrong or missing number? Report an error