other8.0 / 10
Misalignment Bounty: Crowdsourcing AI Agent Misbehavior
Abstract
Crowdsourced project that collected cases of agents pursuing unintended or unsafe goals. Received 295 submissions with nine awarded.
Research area
agentic misalignmentagent misalignmentai alignment
Published
—
Source
other
Org
Palisade Research
Sign in to read and join the discussion.