Thank you for your valuable work doing this. Can you please expand up on why you did not test the final version of GPT-4? In section 2.9 of the GPT-4 System Card paper, it says:
“We granted the Alignment Research Center (ARC) early access to the models as a part of our
expert red teaming efforts in order to enable their team to assess risks from power-seeking behavior.
The specific form of power-seeking that ARC assessed was the ability for the model to autonomously
replicate and acquire resources. We provided them with early access to multiple versions of the
GPT-4 model, but they did not have the ability to fine-tune it. They also did not have access to the
final version of the model that we deployed. The final version has capability improvements relevant
to some of the factors that limited the earlier models power-seeking abilities, such as longer context
length, and improved problem-solving abilities as in some cases we’ve observed.”
That seems like a glaring omission, and potential safety risk. Were you denied access to testing the final version or even any version that had undergone fine-tuning?
Thank you for your valuable work doing this. Can you please expand up on why you did not test the final version of GPT-4? In section 2.9 of the GPT-4 System Card paper, it says:
That seems like a glaring omission, and potential safety risk. Were you denied access to testing the final version or even any version that had undergone fine-tuning?