Loading activity…
@vedantmisra
ASI @DeepMind | @OpenAI (Reasoning / Algorithms Lead) | @HubSpot (ML Labs) | Founder/CEO Kemvi (acq HubSpot) | Physics @Columbia | Husband and Dad
AI
The reported MATH-500 score rose from 7.9% to 48.1% with the same sampling settings.
AI
One post agrees with the message behind the metaphor but asks whether people think in height, width and depth.
AI
In an October 28 post, Li recalled work on AlphaCode and Gemini and outlined plans for a break.
AI
A post says Carbon scored 88.2% on SEC-bench Pro and is rolling out first to cybersecurity specialists.
AI
In the announcement, @bneyshabur places Mirendil-Base in the 100B–500B-parameter range and promises independent evaluations.
AI
A post citing Business Insider says Carbon matches Opus 5.5 in coding.
AI
A post calls extensive, thorough and principled empirical work on both approaches “the answer.”
AI
Google announces a 1 million-token output limit, up from 64,000, for its new model. Initial access goes to trusted cyber defenders through the Fairwind Program.
AI
DeepMind researcher Vedant Misra posted a timeline of shifting attitudes toward computers across decades.
AI
DeepMind researcher Vedant Misra tweets a list of ways flagship AI projects render other work obsolete.
AI
Google's new model improves coding accuracy and agentic tasks at reduced cost.
