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

    Neuronpedia

    INDEX
    Explanations

    they

    np_max-act · gemini-2.0-flash

    This neuron fires on relatively rare or domain-specific content words (especially longer, technical or specialized terms), distinguishing them from common function words.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ategor
    -0.08
    emplo
    -0.07
     iterative
    -0.07
    nj
    -0.07
    J
    -0.07
     مب
    -0.07
     गय
    -0.07
    reas
    -0.07
    .lifecycle
    -0.06
     consumption
    -0.06
    POSITIVE LOGITS
    Visit
    0.07
    oron
    0.06
     principalColumn
    0.06
     Solar
    0.06
     aspire
    0.06
    UTE
    0.06
    .requests
    0.06
    .mContext
    0.06
    	TRACE
    0.06
     قرن
    0.06
    Act Density 0.333%

    No Known Activations