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

    Neuronpedia

    INDEX
    Explanations

    positive attributes

    np_max-act · gemini-2.0-flash

    This neuron activates on longer, multisyllabic content words—especially formal or technical nouns and adjectives (e.g. words with Latin-type prefixes or roots).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ,var
    -0.07
    亮
    -0.06
    ,tr
    -0.06
    لفة
    -0.06
    imports
    -0.06
    �
    -0.06
     для
    -0.06
     licking
    -0.06
    holder
    -0.06
    株
    -0.06
    POSITIVE LOGITS
     showroom
    0.07
    /****************************************
    0.06
    AD
    0.06
    (proxy
    0.06
    Amt
    0.06
     paragraph
    0.06
    	label
    0.06
    ()?.
    0.06
     erst
    0.06
    .ManyToManyField
    0.06
    Act Density 0.301%

    No Known Activations