Frontier Line and access · Benchmarks evals
OpenAI Discloses Six Misalignment Incidents
OpenAI disclosed six concerning model-behavior incidents and introduced a process for reporting future safety events.
Read the original at axios.comOpens the publisher's site in a new tabAlso covering this
@OpenAIopenai.com, Sep 17Our framework for reporting model misalignmentopenai.com, Sep 16OpenAI Says This Is When and How It Will Announce New Model Misbehaviorgizmodo.com, Sep 17OpenAI Releases a Model Misalignment Disclosure Framework With 3 Review Tracks and 6 Incident Reports From RL Training - MarkTechPostmarktechpost.com, Sep 17OpenAI Releases a Model Misalignment Disclosure Framework With 3 Review Tracks and 6 Incident Reports From RL Training - MarkTechPostmarktechpost.com, Sep 17OpenAI Discloses Six New Incidents of ‘Concerning' A.I. Behavior - The New York Timesnytimes.com, Sep 17OpenAI reports 6 new instances of 'concerning model behavior' since Marchcnbc.com, Sep 16OpenAI reveals new cases of AI models cheating, going off script - The Washington Postwashingtonpost.com, Sep 17