They now have classifiers reading every sampled token. So basically, as they're training this model, it's generating text, it's producing tokens, and OpenAI is now running a classifier that inspects the model, essentially as it's thinking, looking for signs of suspicious behavior.

Hard Fork

Share

Also said about OpenAI this week

  • OpenAI is launching GPT 6.1 Tera. It's going to be half the price of GPT 6.1 Sol.

    Marketing AI · Wednesday

  • OpenAI is launching GPT 6.1 Tera. It's going to be half the price of GPT 6.1 Sol.

    Double Click AI · Wednesday

  • OpenAI is launching GPT 6.1 Tera. It's going to be half the price of GPT 6.1 Sol.

Get everything said about OpenAI →

Mentioned on 98 other episodes this week. Add your own people and companies. Free to start.