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

    Neuronpedia

    INDEX
    Explanations

    Code/logs

    np_max-act · gemini-2.0-flash

    This neuron is attuned to numeric literals, i.e. it highlights tokens representing numbers in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    (Path
    -0.08
    (game
    -0.06
    _fecha
    -0.06
    .or
    -0.06
    -Con
    -0.06
    .green
    -0.06
    (fc
    -0.06
     sede
    -0.06
    .Tests
    -0.06
     contingency
    -0.06
    POSITIVE LOGITS
    bij
    0.07
    θηκε
    0.07
    iž
    0.07
    ุด
    0.06
    Hits
    0.06
    Py
    0.06
     lattice
    0.06
    Một
    0.06
    lit
    0.06
    CHANT
    0.06
    Act Density 0.017%

    No Known Activations