• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
  • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI
    Reaction

    Cogentic is said to produce new results on five open problems

    A user describes a Gemini-based system that sends proof drafts to two adversarial verifiers before accepting them.

    HC
    1 Source, ,

    TLDR

    A user describes Cogentic as a Google Research proof-discovery system whose AI provers work in rounds, drawing on summaries of earlier attempts. The user says it produced new results on five open problems, each checked by domain experts. Another user highlights the verifiers’ clean context, arguing it lets provers try unusual ideas without swaying the checks.

    Combined views

    3.9K

    1 Source, first seen 5h ago

    Combined views

    3.9K

    1 Source, first seen 5h ago

    41 likes
    5h ago
    first seen 5h ago
    41 likes
    14 comments
    30 saves
    2 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Featured Source
    14 comments
    30 saves
    2 reposts

    1 Source

    @hwchase17good pattern from this google paper the verifier starts from a clean context, so the explorer's reasoning can't talk it into a bad proof so the explorer can try wild ideas and the verifier tosses the wrong ones same reason deepagents subagents get their own context5h

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    1 Source

    @hwchase17good pattern from this google paper the verifier starts from a clean context, so the explorer's reasoning can't talk it into a bad proof so the explorer can try wild ideas and the verifier tosses the wrong ones same reason deepagents subagents get their own context5h
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet