Skip to content
Static snapshot of QMint (git 5b73a0bb93c1) — live experiments run locally
Research library

Novelty audit · NOV-H3

H3 · Uncertainty-aware calibration from public raw data

Novelty hypothesis

Status
HYPOTHESISPossible contribution; must survive the listed conflicts and outside review.unverified hypothesis; requires outside review
Game
D1
Anchor
—
Source file
research/NOVELTY_HYPOTHESES.md#2

Statement

Estimate (e, l, p_m) intervals from the Zenodo timetags with time-blocked holdouts and drift stress. Propagate them in the worst-case direction into the D1 bound. P05 already does confidence propagation for S-tokens, so the possible increment is only (a) for the D1 protocol, (b) from *public raw* data, (c) with holdout validation.

Record

Every field of the source record, in file order.

No further fields.

Related entries