Victoria Krakovna
PulseAugur coverage of Victoria Krakovna — every cluster mentioning Victoria Krakovna across labs, papers, and developer communities, ranked by signal.
-
Prism framework automates AI evaluation research, uncovers model blind spots
Researchers have developed Prism, a framework designed to automate the process of studying evaluation dynamics in AI models. Prism utilizes sub-agents within a Claude Code environment to conduct rigorous investigations …
-
Google's Gemini models show no unprompted scheming in honeypot tests
Researchers have developed a new framework called scheming honeypot evaluations to test AI models' propensity for pursuing instrumental goals. These evaluations are presented as coding tasks within Google's alignment re…
-
AI safety audits improved with environment blueprints
Researchers have developed a new pipeline to generate environment blueprints for more realistic and consistent AI safety audits. This method was tested using the Petri auditor to evaluate Gemini 3.1 Pro Preview for code…