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

    Neuronpedia

    INDEX
    Explanations

    happy

    np_max-act · gemini-2.0-flash

    This neuron activates on mentions of “wasp nest” (especially the subword “asp” plus “nest”), i.e. whenever the text refers to removing or dealing with a wasp’s nest.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    words that denote emotional or psychological states, especially mood, hostility, or desire.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
    Inputs
    -0.07
    Columns
    -0.07
     зад
    -0.07
     Bentley
    -0.07
    GAN
    -0.07
     eup
    -0.07
    oper
    -0.07
     pools
    -0.06
    tan
    -0.06
    陷
    -0.06
    POSITIVE LOGITS
     verdienen
    0.07
    яется
    0.06
    /power
    0.06
     آنلاین
    0.06
    اهيم
    0.06
    ेण
    0.06
     ISS
    0.06
    ादन
    0.06
    .Small
    0.06
     collider
    0.06
    Act Density 0.233%

    No Known Activations