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

    Neuronpedia

    INDEX
    Explanations

    friendship

    np_max-act · gemini-2.0-flash

    The neuron activates on tokens conveying emotional warmth or friendly affection (e.g. “warming,” “友情”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    端
    -0.06
    /channel
    -0.06
     Walters
    -0.06
     Lexer
    -0.06
    UIAlertView
    -0.06
    ,user
    -0.06
    editor
    -0.06
    -fields
    -0.06
     Trends
    -0.06
     Hunts
    -0.06
    POSITIVE LOGITS
    müş
    0.07
    ±
    0.06
     Occupational
    0.06
     giản
    0.06
    (ERROR
    0.06
     Brotherhood
    0.06
    .Tasks
    0.06
    谱
    0.06
     comrades
    0.06
     relationship
    0.06
    Act Density 0.013%

    No Known Activations