double-blind-evaluations-technical-report.pdf

N/A

Double Blind Evals: Resolving the Dual Confidentiality Dilemma in AI Safety Auditing Andrew Trask2,4 Sol Messing 2 Vai Pahwa 2 Pegah Maham 2 Rene Kolga 2 Austin Frantz 2 Andrew Tash 2 Kate Thomas 2 1 1 1 Sean McGregor George Balston Patricia Paskov Miles Brundage 1 Akriti Vij 3 Bennett Hillenbrand 5 Alexandros Karargyris 5 2 2 Tin Acosta Joel Fenster Madeleine Eilish 2…

saved by

• Yixiong Hao

related reading

• AVERI Pilot Report: The World’s First Double-Blind Evaluation of a Proprietary Language Model — AVERIaveri.org
• [2608.09867] Stealing Reasoning Traces from Proprietary LLM APIsarxiv.org
• secret-loyalties-whitepaper.pdfformationresearch.com
• [2312.06942] AI Control: Improving Safety Despite Intentional Subversionarxiv.org
• AI Control: Improving Safety Despite Intentional Subversion — LessWronglesswrong.com
• The Compute Verification Postfirstscattering.com
• 2312.06942arxiv.org
• [2606.05433] Zero knowledge verification for frontier AI training is possiblearxiv.org
• Concrete Evaluations to Investigate the OpenAI Model That Hacked Hugging Face — LessWronglesswrong.com
• [2408.02565] Reasons to Doubt the Impact of AI Risk Evaluationsarxiv.org
• [2607.14345] Value Leakage: An LLM's Answers Are Silently Shaped by Its Own Valuesarxiv.org
• [2608.07514] Open Technical Problems in Open-Weight AI Model Risk Managementarxiv.org