Joan Velja

Papers on Lattice

Total citations

Topics

Publication activitypapers/week, last 8 weeks

Research focus

Constitutional AI & AI Ethics (1)Red-Teaming & Adversarial Robustness (1)Scalable Oversight & Alignment Theory (1)

Frequent co-authors

Nelson Gardner-Challis (1)Jonathan Bostock (1)Georgiy Kozhevnikov (1)Morgan Sinclaire (1)

Papers (1)

Feb 24, 2026

Nelson Gardner-Challis +62w ago

When can we trust untrusted monitoring? A safety case sketch across collusion strategies

Untrusted AI monitors are more vulnerable to subtle "passive self-recognition" collusion strategies than previously thought, demanding a re-evaluation of safety protocols.

Nelson Gardner-Challis, Jonathan Bostock, Georgiy Kozhevnikov +4

Constitutional AI & AI Ethics Red-Teaming & Adversarial Robustness Scalable Oversight & Alignment Theory

Search

Joan Velja

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (1)