The firm also announced a new system to track, investigate and disclose cases of models misbehaving, or “misalignment”.
OpenAI reveals six more safety issues and unveils plan to disclose incidents
Search
Have an existing account?
Sign In
