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

    Neuronpedia

    INDEX
    Explanations

    test

    np_max-act · gemini-2.0-flash

    The neuron is essentially dead—it never activates (always outputs zero) on any token.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     nameof
    -0.07
     StreamWriter
    -0.07
     wireType
    -0.07
    HeaderValue
    -0.06
     August
    -0.06
     Yuk
    -0.06
    _logical
    -0.06
    .strings
    -0.06
    >>>>>>>
    -0.06
     Glover
    -0.06
    POSITIVE LOGITS
     skirm
    0.07
     lead
    0.06
    @Test
    0.06
     motivating
    0.06
    ै।↵↵
    0.06
     rang
    0.06
    ymph
    0.06
    hope
    0.06
    .↵↵
    0.06
    чно
    0.06
    Act Density 0.000%

    No Known Activations