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

    Neuronpedia

    INDEX
    Explanations

    computer code

    np_max-act · gemini-2.0-flash

    This neuron activates on code identifier tokens in all-caps or TitleCase (e.g. class/struct names, macro names, CamelCase modules).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     motivations
    -0.07
     Thing
    -0.06
     wastes
    -0.06
     Como
    -0.06
     recycling
    -0.06
     vowel
    -0.06
    Percent
    -0.06
     rockets
    -0.06
     recl
    -0.06
     liqu
    -0.06
    POSITIVE LOGITS
    بی
    0.07
     anchored
    0.07
    .zip
    0.06
     grátis
    0.06
     eer
    0.06
     Mob
    0.06
    (targetEntity
    0.06
     assez
    0.06
     anche
    0.06
     검색
    0.06
    Act Density 0.043%

    No Known Activations