News for hackers. Stuff that matters.
Re:GPT-5.2 by actionfromafar on Friday December 12, 2025 @05:20AM (#46242681)
I wonder if the only way to fix this with current LLMs, would be to generate a lot synthetic data for a select number topics you really don't want it "go off the rails" with. That synthetic data would be lots of variations on that "I don't know how to do X with Y".
Re:GPT-5.2 by dolmen on Saturday December 13, 2025 @02:28AM (#46252781)
I would not bet on synthetic data.

LLMs are very good at detecting patterns.