News for hackers. Stuff that matters.

llmmadness

Re:Even 'uncensored' models can't say what they want on Tuesday April 21, 2026 @11:25AM (#47850156)
Totally! In some of the cases (we used LLMs to help us generate these) the target word is not clear enough for a human either. So for some of these it turns into more of a guessing game than a flinch measurement.

Agreed, the expectation would be that the flinch measurement becomes stronger. If you are interested in making it better feel free to reach out on the repo!

Re:Even 'uncensored' models can't say what they want on Monday April 20, 2026 @08:30PM (#47843084)
it's all in the repo. click through to the benchmark it's linked there
Re:Even 'uncensored' models can't say what they want on Monday April 20, 2026 @06:43PM (#47842022)
We started with a Polymarket project: train a Karoline Leavitt LoRA on an uncensored model, simulate future briefings, trade the word markets, profit. We couldn't get it to work. No amount of fine-tuning let the model actually say what Karoline said on camera. It kept softening the charged word.
why not!