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

    Neuronpedia

    INDEX
    Explanations

    in

    np_max-act · gemini-2.0-flash

    The neuron detects code patterns where you look up or compute an element’s index (e.g. calls like `.indexOf`, `.find`, or similar) in a list or string.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     suicidal
    -0.07
     punishable
    -0.06
    ELS
    -0.06
    ุมภาพ
    -0.06
    .prop
    -0.06
     kotlinx
    -0.06
     léč
    -0.06
    аниц
    -0.06
     Крас
    -0.06
     Mark
    -0.06
    POSITIVE LOGITS
    .^
    0.07
     pry
    0.07
     dari
    0.07
    -article
    0.07
    ori
    0.07
    verified
    0.07
     opp
    0.07
     ai
    0.07
    .UI
    0.07
        ↵↵
    0.07
    Act Density 0.024%

    No Known Activations