Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    quest

    np_max-act · gemini-2.0-flash

    This neuron fires on programming and configuration identifiers—terms like class or directory names, code keywords, file names/extensions, and other technical tokens.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    out
    -0.08
     dungeon
    -0.07
    arily
    -0.06
    OUT
    -0.06
     synonym
    -0.06
    ypse
    -0.06
     přid
    -0.06
    [f
    -0.06
    cedures
    -0.05
     coins
    -0.05
    POSITIVE LOGITS
    Crystal
    0.07
     grand
    0.07
    _regularizer
    0.07
    -${
    0.07
    Grand
    0.07
    _precision
    0.07
     Grand
    0.07
     Bast
    0.06
     prac
    0.06
    -cr
    0.06
    Act Density 0.100%

    No Known Activations