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

    Neuronpedia

    INDEX
    Explanations

    Descriptive text

    np_max-act · gemini-2.0-flash

    This neuron activates on placeholder name tokens (e.g. “NAME_#”), signaling the presence of a named entity placeholder.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     �
    -0.07
    	length
    -0.07
    的に
    -0.07
    _pipeline
    -0.07
    �
    -0.07
    _COLLECTION
    -0.07
     Epid
    -0.06
    pixels
    -0.06
    Sch
    -0.06
    .source
    -0.06
    POSITIVE LOGITS
    τικού
    0.06
     *)↵
    0.06
    ISTA
    0.06
    !");
    0.06
    ista
    0.06
     đội
    0.06
    νον
    0.06
     Bers
    0.05
    :h
    0.05
     anak
    0.05
    Act Density 0.416%

    No Known Activations