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

    Neuronpedia

    INDEX
    Explanations

    fonts

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on numeric tokens—especially decimal values and version/build numbers—in the CSS/font‐related code.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    \Product
    -0.07
     triangle
    -0.06
     엘
    -0.06
    Exit
    -0.06
     Sellers
    -0.06
     Redis
    -0.06
     redistributed
    -0.06
    滑
    -0.06
     holdings
    -0.06
     hierarchical
    -0.06
    POSITIVE LOGITS
    (cb
    0.08
    allow
    0.07
    [:
    0.07
    ey
    0.07
    fsp
    0.06
     Gardens
    0.06
     무엇
    0.06
     breve
    0.06
     pł
    0.06
    process
    0.06
    Act Density 0.005%

    No Known Activations