Less is more: UC Berkeley and Google unlock LLM potential through simple sampling

With multiple sampling and self-verification, Gemini 1.5 Pro can outperform o1-preview in reasoning tasks.

New in the last 18 hours

Google will let you make AI podcasts from Gemini’s Deep Research

New in the last 36 hours

Small models as paralegals: LexisNexis distills models to build AI assistant

New in the last 48 hours

Intel’s AI Playground is the perfect ‘everything’ app for newbies—with a catch

Google is now integrating Gemini AI directly into Chrome

New in the last 3 days

Beyond RAG: SEARCH-R1 integrates search engines directly into reasoning models

Meager 8GB of RAM forces Pixel 9a to run “extra extra small” Gemini AI

OpenAI research lead Noam Brown thinks AI ‘reasoning’ models could’ve arrived decades ago

OpenAI’s o1-pro is the company’s most expensive AI model yet