# Long-horizon robot manipulation on CALVIN unseen lang

Source: [Long-horizon robot manipulation on CALVIN unseen lang](<https://www.sota2.com/research/sota/long-horizon-robot-manipulation-on-calvin-unseen-lang>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T06:06:06.645Z

Evaluation of long-horizon robot manipulation capabilities using the CALVIN dataset, specifically focusing on performance in unseen language scenarios. Metrics include sequential task completion rates (1 through 5 tasks) and average trajectory length.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Long-horizon robot manipulation |
| Dataset (including any stated split) | CALVIN unseen lang |
| Metric | Task Completion Rate (1 Task) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Task Completion Rate (1 Task)](<https://www.sota2.com/research/sota/long-horizon-robot-manipulation-on-calvin-unseen-lang.md?metric=Task%20Completion%20Rate%20(1%20Task)>): Higher is better; unit: PERCENT.

- [Task Completion Rate (2 Tasks)](<https://www.sota2.com/research/sota/long-horizon-robot-manipulation-on-calvin-unseen-lang.md?metric=Task%20Completion%20Rate%20(2%20Tasks)>): Higher is better; unit: PERCENT.

- [Task Completion Rate (3 Tasks)](<https://www.sota2.com/research/sota/long-horizon-robot-manipulation-on-calvin-unseen-lang.md?metric=Task%20Completion%20Rate%20(3%20Tasks)>): Higher is better; unit: PERCENT.

- [Task Completion Rate (4 Tasks)](<https://www.sota2.com/research/sota/long-horizon-robot-manipulation-on-calvin-unseen-lang.md?metric=Task%20Completion%20Rate%20(4%20Tasks)>): Higher is better; unit: PERCENT.

- [Task Completion Rate (5 Tasks)](<https://www.sota2.com/research/sota/long-horizon-robot-manipulation-on-calvin-unseen-lang.md?metric=Task%20Completion%20Rate%20(5%20Tasks)>): Higher is better; unit: PERCENT.

- [Avg. Trajectory Length](<https://www.sota2.com/research/sota/long-horizon-robot-manipulation-on-calvin-unseen-lang.md?metric=Avg.%20Trajectory%20Length>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Task Completion Rate (1 Task) | Result date | Report paper |
| --- | --- | --- | --- |
| GR-1 | 76.4 | Not recorded | 68fa71861e209131cea51c7f |
| HULC | 71.5 | Not recorded | 68fa71861e209131cea51c7f |
| MT-R3M | 51.2 | Not recorded | 68fa71861e209131cea51c7f |
| RT-1 | 49.4 | Not recorded | 68fa71861e209131cea51c7f |

### GR-1

Task Completion Rate (1 Task): 76.4. Result ID: 6991eeff3a6c3148490b8611.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71861e209131cea51c7f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71861e209131cea51c7f.md>)

[Implementation](<https://gr1-manipulation.github.io/>)

Recorded evaluation configuration: {"Experiment":"unseen lang"}. This record does not assert independent reproduction.

### HULC

Task Completion Rate (1 Task): 71.5. Result ID: 6991eeff3a6c3148490b860f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71861e209131cea51c7f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d4d1e209131cea38eb6.md>)

Recorded evaluation configuration: {"Experiment":"unseen lang"}. This record does not assert independent reproduction.

### MT-R3M

Task Completion Rate (1 Task): 51.2. Result ID: 6991eeff3a6c3148490b8610.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71861e209131cea51c7f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e011e209131cea3d5b4.md>)

[Implementation](<https://github.com/facebookresearch/r3m>)

Recorded evaluation configuration: {"Experiment":"unseen lang"}. This record does not assert independent reproduction.

### RT-1

Task Completion Rate (1 Task): 49.4. Result ID: 6991eeff3a6c3148490b860e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71861e209131cea51c7f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70411e209131cea49f7d.md>)

[Implementation](<http://github.com/google-research/robotics_transformer>)

Recorded evaluation configuration: {"Experiment":"unseen lang"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
