Integrity report for One LR Doesn't Fit All: Heavy-Tail Guided Layerwise Learning Rates for LLMs

A machine-verified record of the checks Pith has run against this paper: detector runs, findings, signed bundle events, and canonical identifiers.

arXiv:2605.22297 · pith:2026:HHWO7G6NSHDVZVBETWE6DMT7UZ

0Critical

0Advisory

7Detectors run

2026-05-23Last checked

Paper page arXiv integrity.json bundle.json

Detector runs

shingle_duplication completed v0.1.0 · findings 0 · 2026-05-23 17:50:22.239968+00:00

citation_quote_validity completed v0.1.0 · findings 0 · 2026-05-23 15:51:17.860117+00:00

doi_title_agreement completed v1.0.0 · findings 0 · 2026-05-23 08:02:06.975279+00:00

doi_compliance completed v1.0.0 · findings 0 · 2026-05-23 07:53:06.171904+00:00

claim_evidence completed v1.0.0 · findings 0 · 2026-05-23 01:42:59.966890+00:00

cited_work_retraction completed v1.0.0 · findings 0 · 2026-05-22 09:53:12.598486+00:00

ai_meta_artifact skipped v1.0.0 · findings 0 · 2026-05-22 01:33:42.768644+00:00

Findings

No public integrity findings for this paper.

Signed record

The machine-readable record for this paper lives at /pith/HHWO7G6N/integrity.json. Pith Number bundles also include signed pith.integrity.v1 events where a Pith Number exists.