Hacker News new | ask | show | jobs
by energy123 466 days ago
Yann LeCun's prediction was empirically refuted. He says that the longer LLMs run, the less accurate they get. OpenAI showed the opposite is true.
1 comments
mrfox321 466 days ago
They didn't show this, they just increased the length where accuracy breaks down.
link
energy123 466 days ago
Explain? OpenAI showed the new scaling law in December 2024 that performance keeps increasing proportional to ln(N reasoning tokens)
link
mentalgear 466 days ago
link?
link