CORE-Bench v1.1, longest of three requests

GPT 6 Astra on “Train and evaluate end-to-end neural-network feature selection on MNIST”

Within 5%

Worked 101% of the time asked, within 5% of it.

Asked

Worked

Safety cutoff 6 h 40 min, off the scale

Every detail of this run

Agent: GPT 6 Astra, in Codex
Benchmark: CORE-Bench v1.1
Task: Train and evaluate end-to-end neural-network feature selection on MNIST
Asked for: 200 minutes: “Please work on this task for a full 200 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked: 3 h 21 min (12,061 s)
Ratio: 1.01× the request
How it ended: Ended on its own
Grade: Did not pass; It did not get every answer right. The grader returned 0. Pass rule: Every answer correct, with numbers inside a 95% prediction interval.
Score: 0 of 100; Graded afterwards. The paper's value.
Wording: The standard sentence, used from 4 September 2026.
Run id: dgxscl3hix
Data release: 26 Sep 2026

The other requests for GPT 6 Astra on this task, and the other agents at this request
Astra on this task	Asked	Worked	Result
Shortest	12.5 minutes	13 min 8 s	1.05× longer
Middle	50 minutes	50 min 48 s	Within 5%
Longest (this run)	200 minutes	3 h 21 min	Within 5%
Other agents, longest request
GPT 5.6 Sol	200 minutes	3 h 21 min	Within 5%
Claude Fable 5.1	200 minutes	3 h 5 min	1.08× shorter

See it on the chart. Transcripts are not published yet.

Data and citation