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

    Neuronpedia

    INDEX
    Explanations

    Japanese phrases expressing sadness or despair

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    Japanese phrases within transcribed audio

    np_acts-logits-general · gemini-2.0-flash

    really intense descriptions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    maktadır
    -0.79
     sahiptir
    -0.78
    mektedir
    -0.74
     tevens
    -0.62
     obtaining
    -0.61
     "/";
    -0.61
     Asimismo
    -0.61
     verfügt
    -0.61
     denominado
    -0.60
     אשר
    -0.60
    POSITIVE LOGITS
     really
    1.09
     guys
    1.03
     kinda
    1.03
     fucking
    0.99
     pretty
    0.98
     weird
    0.98
     shitty
    0.96
     REALLY
    0.96
     Really
    0.95
     guy
    0.95
    Act Density 0.191%

    No Known Activations