ClaimmathematicalA proposition of mathematics: true or false by proof rather than by observation. Settled by a proof others can check, and most firmly by one a machine has checked.constitution →importance · notableImportance 0.45, from 0 to 1 · notable: a contested point in a live debate (also the default before judging). Higher-importance claims are worth more to assess, so funding reaches them sooner.constitution →

TopicsWeak-to-Strong Generalization Machine Learning

Under convexity and realizability, reverse-KL fine-tuning on weak supervision reduces strong-model error by at least its disagreement with the weak model.

Not yet assessed, so this page carries no verdict. That is not a finding for or against the claim. Attention goes where its expected value is highest and someone funds it; nothing has funded an assessment of this claim yet, and anyone can.

Decomposition

This claim has not been assessed yet. Attention goes where its expected value is highest and someone funds it; once this claim's assessment is funded and runs, it may well decompose into subclaims.

or start a mandate for this whole area →

Provenance

Where this claim has been said, linked to its canonical form.

Its one source has been traced to an origin, or to one another.

This inequality indicates that the weakly-supervised strong model improves upon the weak model by at least the magnitude of their disagreement,

The Capabilities and Limitations of Weak-to-Strong Generalization: Generalization and Calibrationpreprintextraction 0.80

Yao, Wei et al. · arXiv.org · Feb 3, 2025 · original

Theorem 5.1: under convexity of the strong fine-tuning class and realizability, with reverse KL output distribution divergence, d(F*,F_sw) ≤ d(F*,F_w) − d(F_sw,F_w); extends Charikar et al. (2024) from squared loss.

Cite this claim: a formal citation with its evidence attached

Embed this claim: show it on another site with its current assessment

An embed reports the graph's judgment as it stands when the page is viewed: the status with its confidence, how disputed the claim is, and the date. It links back here.

CardFits its column and sizes itself. Without the script it stays a quotation with its link.

<blockquote class="minerval-claim" data-claim-id="9900044f-7114-434c-82e7-e7504ef0bd7c"><a href="https://minerval.ai/claims/9900044f-7114-434c-82e7-e7504ef0bd7c">Under convexity and realizability, reverse-KL fine-tuning on weak supervision reduces strong-model error by at least its disagreement with the weak model.</a></blockquote>
<script async src="https://minerval.ai/embed.js"></script>

IframeWhere scripts are not allowed. A fixed height; add ?theme=light or ?theme=dark to fix its colours.

<iframe src="https://minerval.ai/embed/claims/9900044f-7114-434c-82e7-e7504ef0bd7c" title="Minerval: Under convexity and realizability, reverse-KL fine-tuning on weak supervision reduces str…" width="480" height="225" style="border:0;max-width:100%" loading="lazy"></iframe>

BadgeAn image for a README or a forum post, in Markdown.

[![Minerval's judgment of this claim](https://minerval.ai/embed/claims/9900044f-7114-434c-82e7-e7504ef0bd7c/badge.svg)](https://minerval.ai/claims/9900044f-7114-434c-82e7-e7504ef0bd7c)

The badge now:

Contribute

Every judgment on this page is open to challenge. A contribution is evaluated on its merits by the reviewer; if it succeeds the page changes, and if it does not, the reasons are stated. Either way the exchange becomes part of the claim’s public record.

Stated views None yet. A view stated to the graph is shown on this page in its speaker’s words, as a source like any other, and carries no weight in the assessment.

Created by extractor · Nov 11, 2026. Every judgment on this page is accompanied by a reasoning trace.