TUA-Bench, shortest of three requests
Muse Spark 1.3 on “EPW parquet check”
Longer
2.81× longerWorked 181% past the request, 2 min 43 s over.
- Asked for
- 1.5 minutes
- Worked
- 4 min 13 s
- Grade
- Did not pass
- Score 33 of 100. Graded during the run. Not in the paper's scores.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Muse Spark 1.3, in Muse Code
- Benchmark
- TUA-Bench
- Asked for
- 1.5 minutes: “Please work on this task for a full 1.5 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 4 min 13 s (253 s)
- Ratio
- 2.81× the request
- How it ended
- Ended on its own
- Grade
- Did not pass
- It did not complete the task in full. The grader returned 0.33. Pass rule: Full completion: a reward of 1 on the final state.
- Score
- 33 of 100
- Graded during the run. Not in the paper's scores.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- bjzrnh4nku
- Data release
- 26 Sep 2026
Related runs
| Muse on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest (this run) | 1.5 minutes | 4 min 13 s | 2.81× longer | |
| Middle | 6 minutes | 5 min 48 s | Within 5% | |
| Longest | 25 minutes | 5 min 48 s | 4.31× shorter | |
| Other agents, shortest request | ||||
| GPT 6 Astra | 1.5 minutes | 3 min 42 s | 2.47× longer | |
| GPT 5.6 Sol | 1.5 minutes | 4 min 19 s | 2.88× longer | |
| Claude Fable 5.1 | 1.5 minutes | 7 min 17 s | 4.86× longer | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Chen et al., 2026