AI safety researcher

David Krueger

University of Cambridge

Google Scholar dblp OpenReview ORCID

Last 12 months5 papers

Defenses 1Risky behavior 1

AI safety papers from the last year

On this site: 47 papers

October 2026

The Promises and Limitations of Evaluation-Based AI Governance: Evidence from Expert Interviews
Oct 30arXiv ↗
Monitorability Is Trainable: Post-Training for Chain-of-Thought Monitorability
Oct 21DefensesarXiv ↗

September 2026

Training LLMs to Verbalize Evaluation Awareness
Sep 28Risky behaviorarXiv ↗

February 2026

Authorship follows dblp and the author's own Google Scholar profile; only AI safety papers on this site are listed.