LodeHQ
AI news for people who shipFree, by email

AI · arxiv.org · Wednesday, 1 July 2026

Perplexity differencing cracks hidden finetuning goals in model organisms

A simple perplexity-difference test can expose the hidden finetuning objectives of public model organisms, from backdoors to fabricated facts. Tested on 76 models up to 70B parameters, the technique ranks completions that reveal illicit behavior, achieving state‑of‑the‑art detection on the AuditBench benchmark.

Read the source: arxiv.org ↗In the 1 July issue →