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

    Neuronpedia

    INDEX
    Explanations

    ramp

    np_max-act · gemini-2.0-flash

    The neuron highlights single-word concrete object nouns (tangible things like “ramp,” “dog,” “rose,” “tube,” “safe,” etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    cart
    -0.07
     healing
    -0.07
     counters
    -0.06
     fus
    -0.06
    üler
    -0.06
    арам
    -0.06
    	System
    -0.06
     affects
    -0.06
    upt
    -0.06
    alling
    -0.06
    POSITIVE LOGITS
     �
    0.07
    (schema
    0.06
    /pol
    0.06
     STM
    0.06
    ）。↵↵
    0.06
    Não
    0.06
    sphere
    0.06
    _modifier
    0.06
    кот
    0.06
     strokeLine
    0.06
    Act Density 0.071%

    No Known Activations