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

    Neuronpedia

    INDEX
    Explanations

    the

    np_max-act · gemini-2.0-flash

    The neuron fires on words and phrases used to judge factual consistency—e.g. “fact,” “factually,” “consistent,” and “inconsistent.”

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    phrases indicating the action of requesting or asking someone to do something.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .LA
    -0.07
    _y
    -0.07
    ρία
    -0.07
     pog
    -0.06
    eryl
    -0.06
    网址
    -0.06
     slam
    -0.06
     Commons
    -0.06
     Penalty
    -0.06
     goog
    -0.06
    POSITIVE LOGITS
     marched
    0.06
    {lng
    0.06
     nominated
    0.06
    ньої
    0.06
    าหาร
    0.06
     mạch
    0.06
     castle
    0.06
     wildfire
    0.05
    .scalajs
    0.05
    .series
    0.05
    Act Density 0.020%

    No Known Activations