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

    Neuronpedia

    INDEX
    Explanations

    conversational text

    np_max-act · gemini-2.0-flash

    The neuron fires on “like”-based comparisons (similes) and related phrasal uses of “up” (e.g. “built like a tank,” “things are starting to look up”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Celt
    -0.08
     Oc
    -0.07
    -bo
    -0.07
     utilizing
    -0.06
     مح
    -0.06
     physicist
    -0.06
     Вели
    -0.06
     edilmiş
    -0.06
    	tv
    -0.06
    templates
    -0.06
    POSITIVE LOGITS
     Seller
    0.07
    inea
    0.06
     defective
    0.06
     又
    0.06
     gastrointestinal
    0.06
    .timeline
    0.06
    _peng
    0.06
     shave
    0.06
     Giovanni
    0.06
     개
    0.06
    Act Density 0.148%

    No Known Activations