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

    Neuronpedia

    INDEX
    Explanations

    decision-making

    np_max-act · gemini-2.0-flash

    The neuron selectively fires on language of empowerment or potential—especially the phrase “power to create” and similar “power to …” constructions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Mad
    -0.07
    .de
    -0.06
    vasive
    -0.06
     tsunami
    -0.06
    BitConverter
    -0.06
     BAD
    -0.06
     TH
    -0.06
    gable
    -0.06
    arial
    -0.06
     Madame
    -0.06
    POSITIVE LOGITS
     �
    0.07
     nouvelle
    0.07
    .chdir
    0.06
    _else
    0.06
    :len
    0.06
     originated
    0.06
    >User
    0.06
     Pen
    0.06
    .ToTable
    0.06
     Salv
    0.06
    Act Density 0.128%

    No Known Activations