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

    Neuronpedia

    INDEX
    Explanations

    nd

    np_max-act · gemini-2.0-flash

    This neuron detects ordinal number suffixes (like “st,” “nd,” “rd,” “th”) attached to numerals.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     recreation
    -0.07
    기술
    -0.07
     thirteen
    -0.07
     radio
    -0.07
     taper
    -0.07
     wrongdoing
    -0.06
     belki
    -0.06
     titanium
    -0.06
     twelve
    -0.06
     eighteen
    -0.06
    POSITIVE LOGITS
    st
    0.10
    ST
    0.09
    rd
    0.09
    nd
    0.08
    std
    0.08
    2
    0.07
     hrs
    0.07
     as
    0.07
    MS
    0.07
     yrs
    0.07
    Act Density 0.011%

    No Known Activations