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

    Neuronpedia

    INDEX
    Explanations

    -

    np_max-act · gemini-2.0-flash

    The neuron fires on numeric tokens—especially signed or decimal numbers—i.e. it detects number‐like entries in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     groß
    -0.07
    課
    -0.07
    ίο
    -0.06
     işte
    -0.06
     різних
    -0.06
    γχ
    -0.06
     nhờ
    -0.06
    height
    -0.06
     breakfast
    -0.06
     testimon
    -0.06
    POSITIVE LOGITS
     trans
    0.07
    buquerque
    0.06
     Taco
    0.06
     delic
    0.06
     reins
    0.06
     Crab
    0.06
     NSMutable
    0.06
     slur
    0.06
     invitations
    0.06
     taco
    0.06
    Act Density 0.004%

    No Known Activations