Showing cs.AIShow all
2 papers · 1 filter
cs.AI2026
The Circuit Basis of Metacognitive Failure in Language Models: Interference and Hallucination
Qiyao Liang, Risto Miikkulainen, Ila Fiete
Language models often answer with high confidence, as measured by output entropy, even when they are wrong. We study two situations that lead to such errors, knowledge conflict, wh…
cs.AI2024
How Diffusion Models Learn to Factorize and Compose
Qiyao Liang, Ziming Liu, Mitchell Ostrow +1
Diffusion models are capable of generating photo-realistic images that combine elements which likely do not appear together in the training set, demonstrating the ability to \texti…