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

    Neuronpedia

    INDEX
    Explanations

    genomics

    np_max-act · gemini-2.0-flash

    This neuron activates on numerical tokens (digits and decimal numbers), especially when they appear as numeric values or coordinates.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Own
    -0.07
    Kit
    -0.07
     Subway
    -0.07
    vented
    -0.07
     Vote
    -0.07
     Bread
    -0.06
    aturdays
    -0.06
    chemes
    -0.06
    _opcode
    -0.06
     gastr
    -0.06
    POSITIVE LOGITS
    (sequence
    0.07
     hg
    0.07
     hl
    0.06
    [::-
    0.06
     <?
    0.06
    otope
    0.06
    INESS
    0.06
     drought
    0.06
    ([
↵
    0.06
     vortex
    0.06
    Act Density 0.003%

    No Known Activations