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

    Neuronpedia

    INDEX
    Explanations

    punctuation

    np_max-act · gemini-2.0-flash

    role or section labels that precede responses in formatted dialogues, such as speaker/mode tags and bracketed headers.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     pokoj
    -0.07
     Restore
    -0.07
     майбут
    -0.06
    GameObjectWithTag
    -0.06
    Paper
    -0.06
    Release
    -0.06
     Gorgeous
    -0.06
    their
    -0.06
     IX
    -0.06
    .Tele
    -0.06
    POSITIVE LOGITS
     emb
    0.07
     حي
    0.07
    无码
    0.07
    "))
↵
    0.07
     ={↵
    0.06
    ++)↵
    0.06
    大学
    0.06
    aging
    0.06
    อบ
    0.06
     heatmap
    0.06
    Act Density 0.030%

    No Known Activations