Playing with Vision Embeddings

(prestonbjensen.com)

95 points | by prestoj 3 days ago ago

8 comments

  • joaquincabezas 5 minutes ago

    This article is very well structured and provides just the right amount of details for non-practitioners to enjoy it.

    Mechanistic interpretability is a fun topic to "play with" (good title there). I recommend watching videos featuring Neel Nanda or Chris Olah

  • markusMB 4 hours ago

    Beautiful illustrations I find, 'Playing' is just the free and motivated version of 'exploration'.

    One thought on your nicely illustrated "key observation [is] that neural networks tend to place features along directions": my guess is that the neural net was TOLD to behave that way by choosing e.g. Cosine Loss?

  • agentbraker 24 minutes ago

    Awesome project! Preserving and sharing knowledge like this is incredibly valuable. Thanks for making these resources accessible to everyone.

  • RealityVoid 4 hours ago

    For some reason, the uncanniness of the feature pictures are deeply unsettling for me. It just stirs intense unease. A bit amusing, to be honest.

  • archermarks 2 hours ago

    Nice article! The generated images make me so nostalgic for the early days of AI image generation. DeepDream and others had such uncanny, interesting generations.

  • jcattle 6 hours ago

    Very nice visualizations, thanks for that!

    One thing I still struggle with in my head is how these vision embeddings can then be used to give LLMs eyes.

    Because you somehow need a giant training set which describes images in natural language, no? Is that actually how it works, or is there some smart trick so you don't need to pay labellers a bunch of money to look at pictures and describe them.

    • dilyevsky 5 hours ago

      > Because you somehow need a giant training set which describes images in natural language, no?

      That's definitely one way - they train a text encoder together with an image encoder on a labelled set of images. WL & 3b1b made a nice video on it: https://www.youtube.com/watch?v=iv-5mZ_9CPY

      • jcattle 5 hours ago

        Thanks I'll check out that video