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

    Neuronpedia

    INDEX
    Explanations

    instead

    np_max-act · gemini-2.0-flash

    mentions of numeric values (e.g., years, counts, ratios, or numbered options), often in tabular or reference-style formats.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     projection
    -0.07
    	cfg
    -0.07
     bath
    -0.07
     Dia
    -0.07
     Agency
    -0.07
    Package
    -0.07
    _rp
    -0.06
    _error
    -0.06
    .quick
    -0.06
     diabetes
    -0.06
    POSITIVE LOGITS
    ованих
    0.08
     F
    0.07
    shire
    0.06
     ditch
    0.06
     vida
    0.06
    istle
    0.06
    ilinx
    0.06
    grese
    0.06
    recht
    0.06
     нес
    0.06
    Act Density 0.447%

    No Known Activations