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

    Neuronpedia

    INDEX
    Explanations

    to

    np_max-act · gemini-2.0-flash

    The neuron fires on tokens in the “forgotten to” construction—that is, it detects when the text says someone forgot to do something.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Minuten
    -0.07
    orem
    -0.06
    olu
    -0.06
    ฟอร
    -0.06
    [next
    -0.06
    อดภ
    -0.06
     FIR
    -0.06
    лет
    -0.06
    Tot
    -0.06
     �
    -0.05
    POSITIVE LOGITS
     погод
    0.07
     Highlights
    0.07
    ()].
    0.07
    :.
    0.07
    )!
    0.07
    .Out
    0.06
     forgot
    0.06
     TS
    0.06
    MemoryWarning
    0.06
    !:
    0.06
    Act Density 0.017%

    No Known Activations