WRSWorld Research StateContribute
WORLD RESEARCH STATE

Wen and colleagues with Claude: automated weak-to-strong supervision research

Reported result · significance if the exact claim is established

If verified · conditional significanceClaim scope and evidenceReported result · not independently verified.
If verified · reference score20,000Editorial revision range 7,000–60,000. Not a statistical interval. Included in rankings as an unverified reported result; excluded from achieved-science progress totals.

Provisional assessment of the stated contribution scope. Scientific evidence and contributor attribution are separate.

Scientific record data (JSON) · Challenge or correct a rating

What is claimed?

Ran parallel research agents on weak-to-strong supervision, obtaining and testing combinations of representation-based confidence filtering, evolutionary adapter refinement and posterior-label noise modeling, with selected method transfer to new math and coding datasets. Reported result; not independently verified by WRS. Ranked using a provisional grade and reference estimate for the exact claimed result.

Scientific scope

Repeated access makes the principal evaluation set effectively validation, and agents found shortcuts and evaluation exploits. Independent-domain transfer varies, and a production-scale posterior-label trial showed only a noise-floor improvement. Recovery on selected binary tasks is not a general solution to weak-to-strong alignment or superiority over unrestricted human research.

Evidence state

The source reports this result. WRS has not independently reconstructed the complete argument or validated the reported evidence. The grade and score here express conditional significance; they do not certify correctness. The provisional value enters the main reference rankings with an unverified label, but not achieved-science progress totals.

Full scientific record · Challenge or correct this assessment