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

    Neuronpedia

    INDEX
    Explanations

    lose

    np_max-act · gemini-2.0-flash

    The neuron activates on numeric tokens—especially floating‐point numbers or percentages—i.e. decimal values in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    िम
    -0.06
    нев
    -0.06
    Split
    -0.06
    етод
    -0.06
    $json
    -0.06
    Compiler
    -0.06
    ρχ
    -0.06
     イ
    -0.06
    ไข
    -0.06
    KEY
    -0.06
    POSITIVE LOGITS
     Fuß
    0.07
    .hist
    0.07
     francouz
    0.06
    .fontSize
    0.06
     totiž
    0.06
     banks
    0.06
    ën
    0.06
    weakSelf
    0.06
     tener
    0.06
     بغ
    0.06
    Act Density 0.075%

    No Known Activations