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

    Neuronpedia

    INDEX
    Explanations

    essays

    np_max-act · gemini-2.0-flash

    This neuron activates on blank or whitespace-only tokens, i.e. indentations or empty lines.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Chapel
    -0.07
    .epsilon
    -0.06
    Ha
    -0.06
     Toll
    -0.06
     ']
    -0.06
     Lep
    -0.06
    ec
    -0.06
    (region
    -0.06
    	pass
    -0.06
     Mek
    -0.06
    POSITIVE LOGITS
    .Exists
    0.07
    '){
    0.07
    '][$
    0.07
     nghiên
    0.07
     rune
    0.07
              
↵
    0.06
    llvm
    0.06
    avras
    0.06
    iamond
    0.06
     وأن
    0.06
    Act Density 0.015%

    No Known Activations