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

    Neuronpedia

    INDEX
    Explanations

    value

    np_max-act · gemini-2.0-flash

    This neuron detects object‐property key names (like id, name, value, label, object) in code or JSON.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    cube
    -0.07
    чат
    -0.06
     renovation
    -0.06
     dilig
    -0.06
     ----------------------------------------------------------------------------
    -0.06
    (IL
    -0.06
    contained
    -0.06
    _for
    -0.06
     bereits
    -0.06
     DX
    -0.06
    POSITIVE LOGITS
    ině
    0.07
    ''↵
    0.07
    ATTR
    0.07
    чины
    0.06
    ekli
    0.06
    套
    0.06
     recruiter
    0.06
     Issues
    0.06
     pří
    0.06
     batter
    0.06
    Act Density 0.023%

    No Known Activations