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

    Neuronpedia

    INDEX
    Explanations

    blog posts

    np_max-act · gemini-2.0-flash

    chat-formatting header markers that indicate the start of an assistant response.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     pupils
    -0.07
     Ayrıca
    -0.07
     математи
    -0.07
     měla
    -0.06
     zku
    -0.06
    。この
    -0.06
     colabor
    -0.06
     learning
    -0.06
     sie
    -0.06
     hurdle
    -0.06
    POSITIVE LOGITS
    (Web
    0.07
     Sacred
    0.07
    グ
    0.06
    __;↵
    0.06
     Destination
    0.06
    .post
    0.06
    �
    0.06
    Picture
    0.06
     перш
    0.06
    _dot
    0.06
    Act Density 0.179%

    No Known Activations