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

    Neuronpedia

    INDEX
    Explanations

    how

    np_max-act · gemini-2.0-flash

    phrases that frame a question — especially interrogative words/phrases like "how", "do you", "is", etc.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
     Xin
    -0.06
    ряд
    -0.06
     scooter
    -0.06
     над
    -0.06
     Sticky
    -0.06
    放
    -0.06
     setUp
    -0.06
    145
    -0.05
    ;}
    -0.05
    .resume
    -0.05
    POSITIVE LOGITS
     rgb
    0.07
    Interrupt
    0.06
    =np
    0.06
    @register
    0.06
     KeyboardInterrupt
    0.06
     wir
    0.06
     návr
    0.06
    ligt
    0.06
    ouncements
    0.06
     EVT
    0.06
    Act Density 0.054%

    No Known Activations