Good question, the Base GPT 4.1 no CoT is quite bad (from 66.6% side task success with CoT to 27.2% without), so I’d be surprised if finetunes are different, but I can try.
My guess is that it’s mostly elicitation (the model doesn’t try as hard to insert a backdoor). I think APPS backdoors are not that hard to design.
My guess is that it’s mostly elicitation (the model doesn’t try as hard to insert a backdoor). I think APPS backdoors are not that hard to design.