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

    Kimi model met 93% of unseen power-transformer specs, team says

    A team says it used reinforcement learning to teach a Kimi base model to design medium-power transformers, helping compress multi-week engineering efforts into a few minutes of inference.

    T(
    YC
    GT
    3 Sources, ,

    TLDR

    A team reports training a Kimi base model with reinforcement learning to design medium-power electrical transformers. It says the model met 93% of unseen specifications and helped compress multi-week engineering efforts into a few minutes of inference—the process of running a trained model.

    Combined views

    36.9K

    3 Sources, first seen 23d ago

    Combined views

    36.9K

    3 Sources, first seen 23d ago

    402 likes
    23d ago
    first seen 23d ago
    402 likes
    16 comments
    206 saves
    53 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    16 comments
    206 saves
    53 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    3 Sources

    @gentrajectoryAI is now improving its own model architectures and chips. The power infrastructure it runs on deserves the same attention. We used RL to teach a Kimi base model to design medium-power transformers. It met 93% of unseen specs, helping to compress multi-week engineering efforts into a few minutes of inference. 1/5
    @ycombinatorRT @gentrajectory: AI is now improving its own model architectures and chips. The power infrastructure it runs on deserves the same attenti…
    @teortaxesTexKimi K2.6 btw, not K3 Relatively weak models can go very far with niche RL

    3 Sources

    @gentrajectoryAI is now improving its own model architectures and chips. The power infrastructure it runs on deserves the same attention. We used RL to teach a Kimi base model to design medium-power transformers. It met 93% of unseen specs, helping to compress multi-week engineering efforts into a few minutes of inference. 1/5
    @ycombinatorRT @gentrajectory: AI is now improving its own model architectures and chips. The power infrastructure it runs on deserves the same attenti…
    @teortaxesTexKimi K2.6 btw, not K3 Relatively weak models can go very far with niche RL