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

    Neuronpedia

    INDEX
    Explanations

    Codes

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on isolated single-letter tokens—particularly the column‐heading letters (e.g. “g”, “i”, “c”, “b”) in tabular species listings.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Av
    -0.08
    .AlertDialog
    -0.07
     conception
    -0.07
    OLA
    -0.07
    ione
    -0.06
    اسم
    -0.06
    uales
    -0.06
     }).
    -0.06
     Bard
    -0.06
    -selling
    -0.06
    POSITIVE LOGITS
     업데이트
    0.06
     گوش
    0.06
     přest
    0.06
     wrists
    0.06
     ii
    0.06
    Clear
    0.06
    thumb
    0.06
    ornecedor
    0.06
    —we
    0.06
     [,
    0.06
    Act Density 0.020%

    No Known Activations