James Dao

Hey folks! Pleased to announce we have a new open position for a Founding Research Engineer at Harmony Intelligence. You’ll be responsible for measuring and identifying dangerous AI capabilities across various domains: cybersecurity, biosecurity, persuasion and manipulation, self-exfiltration and self-replication, and more. Location: We're a small but quickly growing remote-first...

Please check out our notebook for figure recreation and to examine your own model for clean-up behavior. Produced as part of ARENA 2.0 and the SERI ML Alignment Theory Scholars Program - Spring 2023 Cohort Fig 5: Correlation between DLA of writer head and DLA of [clean-up heads output dependent...