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

    Neuronpedia

    INDEX
    Explanations

    Code variables

    np_max-act · gemini-2.0-flash

    The neuron responds to numeric literal tokens—especially floating‐point constants—in source code.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .spacing
    -0.07
    ную
    -0.07
     giants
    -0.07
     fences
    -0.06
    ":"+
    -0.06
    deş
    -0.06
     Tome
    -0.06
    TEGER
    -0.06
    NON
    -0.06
     schöne
    -0.06
    POSITIVE LOGITS
    ieren
    0.07
     Utf
    0.06
     Stein
    0.06
     Abs
    0.06
     ziyaret
    0.06
     δυ
    0.06
    Adjacent
    0.06
     McD
    0.06
     vb
    0.06
     ControllerBase
    0.06
    Act Density 0.008%

    No Known Activations