spesrobotics/wire_pick_place_multi_view, measured from the public release on 13 September 2026. frame level, every row read
Nobody opens a robot dataset and watches all of it. We read every frame of every episode and report what is in there.
This is the release's own footage for one of the twelve episodes we flagged. The gripper command is the same in the first frame and the last, so the wires are never touched. The twelve are the last twelve episodes in the release, one after another, which is what a recording session looks like when it is left running.
episode 201 of spesrobotics/wire_pick_place_multi_view, cut from the release's own video, 18 seconds. Nothing is re-rendered or sped up.
We read every frame of spesrobotics/wire_pick_place_multi_view, 213 episodes and 229,014 rows. In 12 of them the gripper command never changes from the first frame to the last, so there is no grasp in those episodes at all. On a manipulation corpus that is the one episode type a policy cannot learn the task from.
| property | value |
|---|---|
| episodes | 213 |
| frames | 229,014 |
| frame rate | 60 fps |
| robot | xarm_end_effector |
| action dimensions | 7 |
| cameras | 2 |
| distinct task strings | |
| shortest episode | not declared |
| median episode | not declared |
| longest episode | not declared |
A trailing hold is the number of frames at the end during which the commanded action does not change by more than 1e-4 on any component. Seconds come from the frame rate the release declares.
| trailing hold at least | episodes | share |
|---|---|---|
| 1 frames , 0.02 seconds | 213 | 100.0% |
| 5 frames , 0.08 seconds | 213 | 100.0% |
| 10 frames , 0.17 seconds | 213 | 100.0% |
| 30 frames , 0.5 seconds | 211 | 99.1% |
| 60 frames , 1.0 seconds | 211 | 99.1% |
The release is a meta directory plus one or more parquet files. Read data/chunk-000/file-000.parquet, group rows by the episode ranges in meta/episodes, and compute the same columns. No credentials.
This page is what a published release will tell you on its own. The full grading removes the part that scores worst and retrains on what is left, then reports whether the smaller corpus scores higher, with rollout counts and intervals. The one result measured on physical hardware is Stera-10M, two arms on a four arm SO-ARM101 bench, 21.9% against 36.9% task success across 1,270 rollouts.