Anonymous
8/6/2025, 2:54:04 AM
No.106157461
>>106157178
>according to policy #13
Does the model actually have a numbered list of refusal policies baked in? I wonder if you could extract them one at a time by prefilling "<think>According to policy #N, ..." and see what it says
>according to policy #13
Does the model actually have a numbered list of refusal policies baked in? I wonder if you could extract them one at a time by prefilling "<think>According to policy #N, ..." and see what it says