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

    Neuronpedia

    INDEX
    Explanations

    Legal/political processes

    np_max-act · gemini-2.0-flash

    The neuron responds to mentions of the game “token” count—in particular words about losing tokens (e.g. “lose,” “all,” “five,” “tokens”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     userData
    -0.07
    Showing
    -0.07
    这一
    -0.06
    Discuss
    -0.06
    rowable
    -0.06
     ноя
    -0.06
     Smartphone
    -0.06
    、そう
    -0.06
    [*
    -0.06
    這
    -0.06
    POSITIVE LOGITS
     lstm
    0.07
    XY
    0.07
     hips
    0.07
     didnt
    0.06
     Swing
    0.06
     tilted
    0.06
    ения
    0.06
     перер
    0.06
     factor
    0.06
     Spells
    0.06
    Act Density 0.164%

    No Known Activations