Comment on Little Bobby Tables has a baby sister. Meet Sally Ignore Previous Instructions.
minorninth@lemmy.world 8 months agoGPT-3.5 seems to have a problem of recency bias. With long enough input it can forget its prompt or be convinced by new arguments.
GPT-4 is not immune though better.
I’ve had some luck with a post-prompt. Put the user’s input, then follow up with a final sentence reminding the model of the prompt and desired output format.
vcmj@programming.dev 8 months ago
Yes, that’s by design, the networks work on transcripts per input, it does genuinely get cut off eventually, usually it purges an entire older line when the tokens exceed a limit.
minorninth@lemmy.world 7 months ago
I’m talking about using the ChatGPT API to make a chat bot. Even when the user’s input is just one sentence, it can cause ChatGPT to forget its prompt.
vcmj@programming.dev 7 months ago
Ah, even then it could just be a consequence of training samples usually being chronological so it learns to think that way. I did find the pattern also applies to GPT trained on long articles where you’d expect it not to, so wanted to just explain why that might be.
vcmj@programming.dev 8 months ago
Or I should explain better: most training samples will be cut off at the top, so the network sort of learns to ignore it a bit.