‹ BackHN Continuity

Thread

Cache-to-Cache: Direct Semantic Communication Between LLMs (2025)

109 points · 22 comments · rochansinha

  1. cubefox · · focus · HN ↗
    So the models will not only be using more and more Neuralese in their CoT (like GPT-6), but different agents will also be able to communicate with each other in Neuralese. It's not looking good for monitorability.
    1. sparky_twofort · · focus · HN ↗
      Is Neuralese in no way decodable into a human-interpretable system? Genuine question -- I don't know the answer.
      1. TeMPOraL · · focus · HN ↗
        Probably not without a sufficiently powerful LLM from same family, or something equivalent, to act as a translation layer, which includes the risk of the translator lying to you.
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.