LLMs can't self-correct in reasoning tasks, DeepMind study finds (bdtechtalks.com) | HN Screenshots

YHacker Newsscreenshots

Screenshot of bdtechtalks.com: LLMs can't self-correct in reasoning tasks, DeepMind study finds

First screen, captured 3 years ago.

Looks similar

Screenshot of technologyreview.com: Don't be fooled–LLMs don't reason

Don't be fooled–LLMs don't reason (technologyreview.com)

41 points by leopoldj 24 days ago | 42 comments

MIT Technology Review opinion piece by Thore Graepel arguing that LLMs don't truly reason, ten years after AlphaGo's match against Lee Sedol.

llms ai-reasoning alphago deepmind

Screenshot of eugeneyan.com: Task-specific LLM evals that do and don't work

Task-specific LLM evals that do and don't work (eugeneyan.com)

171 points by ZeljkoS 1 year ago | 41 comments

Screenshot of eugeneyan.com: Don't mock machine learning models in unit tests

Don't mock machine learning models in unit tests (eugeneyan.com)

55 points by 7d7n 2 years ago | 50 comments

Screenshot of technologyreview.com: AI recursive self-improvement might not come so quickly after all

AI recursive self-improvement might not come so quickly after all (technologyreview.com)

48 points by dgellow 1 month ago | 45 comments

Screenshot of ibm.com: The discovery that changed how scientists think about memory

The discovery that changed how scientists think about memory (ibm.com)

105 points by rbanffy 4 months ago | 42 comments

Screenshot of techxplore.com: “Learning how to Learn” will be next generation's most needed skill

“Learning how to Learn” will be next generation's most needed skill (techxplore.com)

82 points by Brajeshwar 1 year ago | 54 comments