Three LLM judges couldn't tell an AI agent's real memory from a counterfeit one(kairos-experiment.com)
5 pointsby gcolella2 hours ago1 comment