# UnlockPickup on MiniGrid 8x8 (test)

Source: [UnlockPickup on MiniGrid 8x8 (test)](<https://www.sota2.com/research/sota/unlockpickup-on-minigrid-8x8-test>)

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

Data updated: 2026-06-08T06:42:06.023Z

Evaluation of agent performance on the UnlockPickup task within the 8x8 configuration of the MiniGrid environment, using the test split. Metrics include Success Rate, Average Steps to Goal, and various reward measures.

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 | UnlockPickup |
| Dataset (including any stated split) | MiniGrid UnlockPickup 8x8 (test) |
| Metric | Reward AUC |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [Reward AUC](<https://www.sota2.com/research/sota/unlockpickup-on-minigrid-8x8-test.md?metric=Reward%20AUC>): Higher is better; unit: SCALAR.

- [Success Rate](<https://www.sota2.com/research/sota/unlockpickup-on-minigrid-8x8-test.md?metric=Success%20Rate>): Higher is better; unit: PERCENT.

- [Avg Steps to Goal](<https://www.sota2.com/research/sota/unlockpickup-on-minigrid-8x8-test.md?metric=Avg%20Steps%20to%20Goal>): Lower is better; unit: SCALAR.

- [Avg Reward](<https://www.sota2.com/research/sota/unlockpickup-on-minigrid-8x8-test.md?metric=Avg%20Reward>): Higher is better; unit: SCALAR.

- [Total Wins](<https://www.sota2.com/research/sota/unlockpickup-on-minigrid-8x8-test.md?metric=Total%20Wins>): Higher is better; unit: SCALAR.

- [Total Steps](<https://www.sota2.com/research/sota/unlockpickup-on-minigrid-8x8-test.md?metric=Total%20Steps>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Reward AUC | Result date | Report paper |
| --- | --- | --- | --- |
| ULPS | 1886.8 | Not recorded | 6a264c8a90966ade919be9ad |
| Uncalibrated LLM | 1113.77 | Not recorded | 6a264c8a90966ade919be9ad |
| Q-Learning (GRPO) | 627.05 | Not recorded | 6a264c8a90966ade919be9ad |
| DQN (GRPO) | 135.5 | Not recorded | 6a264c8a90966ade919be9ad |
| Linear RL (PPO) | -290 | Not recorded | 6a264c8a90966ade919be9ad |
| Unguided RL (PPO) | -358.18 | Not recorded | 6a264c8a90966ade919be9ad |

### ULPS

Reward AUC: 1886.8. Result ID: 6a2662e6fbc373592c227660.

[Reporting paper](<https://www.sota2.com/research/paper/6a264c8a90966ade919be9ad.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a264c8a90966ade919be9ad.md>)

Recorded evaluation configuration: {"calibration":"uncertainty-aware"}. This record does not assert independent reproduction.

### Uncalibrated LLM

Reward AUC: 1113.77. Result ID: 6a2662e6fbc373592c22765c.

[Reporting paper](<https://www.sota2.com/research/paper/6a264c8a90966ade919be9ad.md>)

Recorded evaluation configuration: {"calibration":"none"}. This record does not assert independent reproduction.

### Q-Learning (GRPO)

Reward AUC: 627.05. Result ID: 6a2662e6fbc373592c22765e.

[Reporting paper](<https://www.sota2.com/research/paper/6a264c8a90966ade919be9ad.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Algorithm":"GRPO"}. This record does not assert independent reproduction.

### DQN (GRPO)

Reward AUC: 135.5. Result ID: 6a2662e6fbc373592c22765f.

[Reporting paper](<https://www.sota2.com/research/paper/6a264c8a90966ade919be9ad.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Algorithm":"GRPO"}. This record does not assert independent reproduction.

### Linear RL (PPO)

Reward AUC: -290. Result ID: 6a2662e6fbc373592c22765b.

[Reporting paper](<https://www.sota2.com/research/paper/6a264c8a90966ade919be9ad.md>)

Recorded evaluation configuration: {"Algorithm":"PPO"}. This record does not assert independent reproduction.

### Unguided RL (PPO)

Reward AUC: -358.18. Result ID: 6a2662e6fbc373592c22765d.

[Reporting paper](<https://www.sota2.com/research/paper/6a264c8a90966ade919be9ad.md>)

Recorded evaluation configuration: {"Algorithm":"PPO"}. This record does not assert independent reproduction.

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