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

    Neuronpedia

    INDEX
    Explanations

    GenerationType

    np_max-act · gemini-2.0-flash

    This neuron specializes in spotting numeric literal tokens—especially decimal/floating-point values—embedded in code or annotations.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    提高
    -0.07
    hs
    -0.07
     Yoshi
    -0.07
    DCF
    -0.07
     sphere
    -0.07
    ali
    -0.07
     має
    -0.06
    łe
    -0.06
     suis
    -0.06
    ôn
    -0.06
    POSITIVE LOGITS
     ;;↵
    0.07
    ////////////////////////////////
    0.06
    (FILE
    0.06
     substantive
    0.06
                                ↵
    0.06
     Birth
    0.06
     Vitamin
    0.06
    ='../
    0.06
    ナ
    0.06
    (bodyParser
    0.06
    Act Density 0.003%

    No Known Activations