9 comments

  • mncharity 7 minutes ago

    > the hidden state for different layers carry meaningful self-awareness signal for various situations.

    Is it plausible to wonder if some developer judgement feels, like maybe "the code I just wrote is clean/crufty", or "things came together smoothly/janky", might have extractable signals in some models?

    If so, might one create a shopping list of desired signals to check for in a model, as with activation steering concepts, where one checks whether and how hard each concept can usefully be nudged?

  • olafura 6 minutes ago

    Was actually pulling on a similar thread as I saw announcement so I integrated it just for fun. Have been only running this on my Framework Desktop but should be runnable elsewhere https://github.com/olafura/gemma-4-mic-transcribe

  • astrobiased 2 hours ago

    Is this in any way similar to Goodfire's work? https://www.goodfire.ai/research/rlfr#

    • HenryNdubuaku 2 hours ago

      Thats an interesting outlook, loosely similar.

  • zdw an hour ago

    Have you benched this for coding tasks, with a fallback to a larger local model, for example Qwen-3.6-27B?

    Or using it for sub-tasks, where a framework with a larger primary model dispatches simpler jobs ("summarize ...", etc.) to it?

  • cacio-e-pepe 7 hours ago

    > So we did mechanistic studies on small models, Gemma 4 particularly, and found the hidden state for different layers carry meaningful self-awareness signal for various situations.

    Neat! Just to make sure I understand - you trained your probe layer to take this hidden state and predict p(wrong)?

    Curious to learn more. Any more info on your approach (esp the mechanistic study)?

    • HenryNdubuaku 5 hours ago

      Correct, the study is verbose, we will compile into a neat shareable report and publish once we solve the pending caveats. Interesting username btw haha.