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

    Neuronpedia

    INDEX
    Explanations

    Totals and amounts

    np_max-act · gemini-2.0-flash

    the presence of numeric or quantitative information (numbers, totals, measurements) or tokens in factual/data-heavy contexts.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
     chiều
    -0.07
    roring
    -0.07
    _srv
    -0.07
     downwards
    -0.06
    Whilst
    -0.06
     спря
    -0.06
    ...'
    -0.06
    POOL
    -0.06
    )+"
    -0.06
    visibility
    -0.06
    POSITIVE LOGITS
     airplanes
    0.07
     Cause
    0.07
    isdiction
    0.07
    >f
    0.06
     Southeast
    0.06
    ickets
    0.06
     countertops
    0.06
     warranted
    0.06
     envisioned
    0.06
     рис
    0.06
    Act Density 0.975%

    No Known Activations