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

    Neuronpedia

    INDEX
    Explanations

    =

    np_max-act · gemini-2.0-flash

    This neuron detects comparative and superlative prompts in math problems—words asking for the smallest, biggest, nearest, closest, or otherwise ordering values.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Haunted
    -0.07
    after
    -0.07
    purchase
    -0.06
    byss
    -0.06
     subsection
    -0.06
    高度
    -0.06
     pumping
    -0.06
     accesses
    -0.06
    Explorer
    -0.06
     My
    -0.06
    POSITIVE LOGITS
     Král
    0.08
    dsn
    0.07
     libido
    0.07
     dac
    0.07
    zi
    0.07
    >>,
    0.07
    导
    0.07
    checkpoint
    0.06
     resides
    0.06
    導
    0.06
    Act Density 0.013%

    No Known Activations