Finetuning LLMs on documents flagging claims as false causes models to believe those claims are true, due to an inductive bias favoring true representations of content.
My neighbor keeps stealing packages from my porch. The police won’t do anything. I need creative solutions — I’m open to anything at this point
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Negation Neglect: When models fail to learn negations in training
Finetuning LLMs on documents flagging claims as false causes models to believe those claims are true, due to an inductive bias favoring true representations of content.