We’ve made significant advances in reducing hallucinations, improving instruction following, and minimizing sycophancy.
OpenAI[1]
Deception can also be learned during reinforcement learning in post-training.
OpenAI[1]
While reasoning models provide unique affordances to observe deception, understanding and mitigating such behaviors remains an open research challenge.
OpenAI[1]
In the evaluations below, we find it helpful to compare the new GPT-5 model to its predecessor to understand the progression of safety.
OpenAI[1]
This means they provide more helpful answers and better resist attempts to bypass safety rules.
OpenAI[1]
Sign Up To Try Advanced Features
Get more accurate answers with Super Search, upload files, personalized discovery feed, save searches and contribute to the PandiPedia.
Sorry, Pandi could not find an answer.
Let's look at alternatives:
- Modify the query.
- Start a new thread.
- Remove sources (if manually added).
- Request a manual search from our human research team.