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

    Neuronpedia

    INDEX
    Explanations

    Summarization/core information

    np_max-act · gemini-2.0-flash

    This neuron detects the “Yes”/“No” answer prompt—specifically the quoted options “Yes” and “No” in the question instruction.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     accidental
    -0.07
     regain
    -0.07
    Ult
    -0.07
    _usec
    -0.07
     extradition
    -0.06
    MED
    -0.06
     Loan
    -0.06
    imus
    -0.06
    evity
    -0.06
    效
    -0.06
    POSITIVE LOGITS
    0.07
    -return
    0.06
     объем
    0.06
    0.06
    (n
    0.06
    >r
    0.06
     comparer
    0.06
    poon
    0.06
     Scalars
    0.06
    abilidad
    0.06
    Act Density 0.011%

    No Known Activations