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

    Neuronpedia

    INDEX
    Explanations

    use

    np_max-act · gemini-2.0-flash

    The neuron activates on imperative action words (e.g. “use,” “add”) in instruction-style prompts.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     baskets
    -0.07
    Jones
    -0.07
    -fold
    -0.07
    fold
    -0.07
    _pf
    -0.07
    beck
    -0.07
    _REPORT
    -0.06
    Deep
    -0.06
    ewolf
    -0.06
    #create
    -0.06
    POSITIVE LOGITS
     تجه
    0.07
    ("!
    0.07
     fecha
    0.06
    _EC
    0.06
     perso
    0.06
     �
    0.06
     alleen
    0.06
     Türkçe
    0.06
     subsidiaries
    0.06
    اصيل
    0.06
    Act Density 0.021%

    No Known Activations