#large-reasoning-models

[ follow ]
#machine-learning #ai-performance #inverse-scaling #ai-alignment #apple-research #cognitive-science
fromTechzine Global
9 months ago

Thinking too long makes AI models dumber

Claude models showed a notable sensitivity to irrelevant information during evaluation, leading to declining accuracy as reasoning length increased. OpenAI's models, in contrast, fixated on familiar problems.
Artificial intelligence
Apple
fromInfoQ
10 months ago

Apple's Illusion of Thinking Paper Explores Limits of Large Reasoning Models

LRMs show a critical collapse threshold in reasoning ability as puzzle complexity increases.
Artificial intelligence
fromZDNET
10 months ago

Apple's 'The Illusion of Thinking' is shocking - but here's what it missed

Apple's report debunks the myth of thinking in Large Reasoning Models, attributing their functioning to advanced pattern matching.
[ Load more ]