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

    Neuronpedia

    INDEX
    Explanations

    alloc

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on numeric literal tokens—especially floating-point constants—in source code.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    тий
    -0.07
    .getStatus
    -0.07
    (styles
    -0.06
     footh
    -0.06
    intree
    -0.06
    зм
    -0.06
    \xf
    -0.06
     исполн
    -0.06
     todas
    -0.06
    })();
    -0.06
    POSITIVE LOGITS
    (proto
    0.07
     近
    0.07
     Slots
    0.07
     having
    0.06
     delaying
    0.06
     knowing
    0.06
     Não
    0.06
    imately
    0.06
    окумент
    0.06
     Canterbury
    0.06
    Act Density 0.001%

    No Known Activations