Vassar Robotics · Reproducible evaluation
Explore completed rollouts while the evaluation runs.
Explore MolmoAct 2 across LIBERO tasks with synchronized model-facing cameras, exact task definitions, simulator success criteria, and content-bound provenance.
In-progress evidence
Completed paper-protocol task shards
This partial snapshot does not claim a final paper result; every rate below uses only completed tasks.
- Provisional four-suite rate
- 97.2%
- 1,944 of 2,000 completed rollouts
- Completed tasks
- 40
- 40 planned across four suites
- Completed rollouts
- 2,000
- Fixed-seed episodes
- Camera videos
- 4,000
- Agent + eye-in-hand
Provisional snapshot captured Aug 16, 2026, 5:10 PM UTC. Coverage and aggregate rates will change as more tasks commit.
Supplemental evaluation
LIBERO-90 stays separate
Any completed LIBERO-90 task shards remain provisional and separate from the four paper-protocol suites.
- Provisional rate
- 36.2%
- 1,285 successes
- Completed rollouts
- 3,550
- 71 of 90 tasks complete
Benchmark suites
Explore by suite
Open a suite to compare task-level rates and inspect every rollout.
LIBERO-10
Ten long-horizon tasks from the MolmoAct 2 paper evaluation protocol.
LIBERO-Goal
Tasks that vary the target goal while retaining familiar scenes and objects.
LIBERO-Object
Tasks centered on manipulating different object categories.
LIBERO-Spatial
Tasks that test spatial relationships and placement precision.
LIBERO-90
A 90-task supplemental evaluation, reported separately from the paper score.
What you are seeing
The model’s actual visual inputs
Each rollout records both 256×256 observations after LIBERO’s image processor and before the policy processor. The H.264 streams are silent and synchronized by control step.
observation.images.imageobservation.images.image2