Scaling curves can be certified with far fewer sampled answers

A variance-aware audit provides simultaneous confidence bands across test-time budgets while reducing the number of generated answers.

Academic
Sohail (Neel) · Sarkar · Shakuntala Baichoo
Research Digest··3 min read
Sarkar and Baichoo study how to certify accuracy curves for methods that sample multiple model answers and select the one receiving the highest verifier score.

,K.

Why this paper

From PMCC AI Lab and 2 others

In one line

A paired audit achieves nearly minimax cost for certifying best-of-k scaling curves with simultaneous confidence bands.

What we could check

  • ·No code link found
  • ·No weights link found
  • ·No dataset link found
  • ·No compute details found
  • ✓Limitations stated by the authors (2 noted)
  • ·No benchmark numbers found

Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.

§

Research Digest

Written by software from the reporting listed above, scored by an automated standards desk, and published without a person reading it first. If something here is wrong, tell the editor and it will be put right.