arXiv (All)
open access
When Do Internal Probes Beat Reading the Answer? Miscalibrated Readouts and Behavior-Concealed Knowledge in Language Models
A 0.6B language model, asked to verify 1,200 logical conclusions (half valid, half corrupted by a single semantic edit), answers YES every time. Judged by behavior it discriminates nothing; linear probes on its hidden states read the correct verdict at 0.96 AUC, transferring to unseen logical struct
Conceptio Open Archive
Analyze & explore related research in Conceptio
Connect this record directly into your research workflow or AI agent. Explore citations across open repositories, query full text, and export structured proof bundles.
This document is indexed with metadata only: full text is not available in the archive for this record. Open the official source ↗
Related documents
- Misrecognition, Misrecognition, and Fallibility#747018
- Beyond a Single Factor: Evidence of Multidimensionality in the ASQoL and Its Implications for Scoring#486076
- Beyond a Single Factor: Evidence of Multidimensionality in the ASQoL and Its Implications for Scoring#483850
- Beyond task performance: morphological branching as a diagnostic metric for polish inflectional language models#270340
- Falsificationism – the incomplete story about choosing good experiments#1089882
Record · ID 927857
Retrieved via Conceptio: every document is proof-bundled with source, license, and retrieval metadata.
Retrieved via Conceptio: every document is proof-bundled with source, license, and retrieval metadata.
Bring this data into your agent: Connect via MCP or API · 20 free credits to start · no card required
Read the Docs → · View CLI source on GitHub · Conceptio Open Knowledge Archive
Read the Docs → · View CLI source on GitHub · Conceptio Open Knowledge Archive