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

    Neuronpedia

    INDEX
    Explanations

    programming-related content

    np_max-act · gemini-2.0-flash

    The neuron activates on auxiliary and modal verbs (need, will, would, should, can, etc.) that frame requests or conditions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Fan
    -0.07
     Leather
    -0.06
     gor
    -0.06
     začí
    -0.06
     malignant
    -0.06
    (Call
    -0.06
     Wilhelm
    -0.06
    .functional
    -0.06
    ToDo
    -0.06
    quires
    -0.06
    POSITIVE LOGITS
    ‑
    0.07
    _sy
    0.07
     снова
    0.06
     거
    0.06
    许
    0.06
    _EDGE
    0.06
     sofa
    0.06
    />.↵
    0.06
     honestly
    0.06
    .aspx
    0.06
    Act Density 0.148%

    No Known Activations