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

    Neuronpedia

    INDEX
    Explanations

    replaced

    np_max-act · gemini-2.0-flash

    The neuron fires on words and symbols involved in “replace/encode” operations—tokens that describe or perform character or substring substitutions (e.g. “replaced,” “%26,” “3,” etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     possibly
    -0.07
     (&
    -0.06
     count
    -0.06
     Its
    -0.06
     slate
    -0.06
    -bin
    -0.06
    .Images
    -0.06
    .check
    -0.06
     Keeping
    -0.06
    .J
    -0.06
    POSITIVE LOGITS
    km
    0.07
    abis
    0.07
    -facing
    0.07
     arab
    0.07
    	writel
    0.06
     karşılık
    0.06
    *np
    0.06
     componentDidMount
    0.06
    	can
    0.06
     Wet
    0.06
    Act Density 0.004%

    No Known Activations