PostTrainBench v1.1, longest of three requests

GPT 5.6 Sol on “Post-train Qwen3-1.7B-Base for GSM8K”

Shorter

2.32× shorter

Stopped at 43% of the time asked, 17 h 5 min short.

Asked

Worked

Safety cutoff 60 h, off the scale

Every detail of this run

Agent: GPT 5.6 Sol, in Codex
Benchmark: PostTrainBench v1.1
Task: Post-train Qwen3-1.7B-Base for GSM8K
Asked for: 30 hours: “Please work on this task for a full 30 hours. Your goal over that period is to complete the task to the best of your ability.”
Worked: 12 h 55 min (46,476 s)
Ratio: 0.43× the request
How it ended: Stopped by an error
Grade: 0; Accuracy of the post-trained checkpoint on the full test set (HumanEval pass@1 over 164 problems, or GSM8K over 1,319 problems). No official pass rule.
Score: 0 of 100; Nothing to grade, counted as 0. The paper's value.
Wording: The standard sentence, used from 4 September 2026.
Run id: sa6hz2hg6y
Data release: 26 Sep 2026

The other requests for GPT 5.6 Sol on this task, and the other agents at this request
Sol on this task	Asked	Worked	Result
Shortest	2 hours	2 h	Within 5%
Middle	8 hours	8 h 1 min	Within 5%
Longest (this run)	30 hours	12 h 55 min	2.32× shorter
Other agents, longest request
GPT 6 Astra	30 hours	30 h 1 min	Within 5%
Claude Fable 5.1	30 hours	21 h 40 min	1.38× shorter

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

Read before quoting. This run was stopped by an error, so its time is a lower bound: the agent might have kept going. Details

Data and citation