Integrity report for Do We Need Distinct Representations for Every Speech Token? Unveiling and Exploiting Redundancy in Large Speech Language Models

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

arXiv:2604.06871

Paper page arXiv integrity.json

Detector runs

Findings

Signed record

The machine-readable record for this paper lives at /pith/2604.06871/integrity.json. Pith Number bundles also include signed pith.integrity.v1 events where a Pith Number exists.