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

    Neuronpedia

    INDEX
    Explanations

    looking down

    np_max-act · gemini-2.0-flash

    The neuron is detecting special control or markup tokens (e.g. header/start-of-text markers) rather than normal words.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Password
    -0.06
    (pop
    -0.06
     neurotrans
    -0.06
    ofday
    -0.06
     ass
    -0.06
    rends
    -0.06
    Summon
    -0.06
     свид
    -0.06
    Jay
    -0.06
     oben
    -0.06
    POSITIVE LOGITS
    /min
    0.07
    ческой
    0.07
    Networking
    0.07
    _shader
    0.06
     implementing
    0.06
     ment
    0.06
     Auss
    0.06
     noe
    0.06
    timeofday
    0.06
     обращ
    0.06
    Act Density 0.006%

    No Known Activations