https://ai.ngchi.online/tags/model-security/ Recent content in Model Security on AI Notes Hugo en-us Wed, 14 Oct 2026 09:08:39 +0700 https://ai.ngchi.online/posts/four-elicitation-methods-failed-the-fifth-guessed-all-256-triggers/ Wed, 14 Oct 2026 09:08:39 +0700 https://ai.ngchi.online/posts/four-elicitation-methods-failed-the-fifth-guessed-all-256-triggers/ Five elicitation methods were run against a model with a chemistry classifier encrypted into its weights. Four returned chance accuracy. The fifth recovered it by enumerating all 256 possible triggers, which the authors say themselves.