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

    Neuronpedia

    INDEX
    Explanations

    model: here

    np_acts-logits-general · gemini-2.5-flash-lite

    assistant-style boilerplate that introduces options or bullet-pointed rewrites/explanations in instructional responses.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     تفصیل
    0.43
    перь
    0.41
    itement
    0.40
     açıklam
    0.39
     panoram
    0.39
     panor
    0.37
     raccont
    0.37
    explique
    0.37
     terakhir
    0.36
     자랑
    0.36
    POSITIVE LOGITS
     here
    0.86
    Here
    0.83
     Here
    0.78
    here
    0.76
     aquí
    0.64
     यहां
    0.62
     aqui
    0.62
     هنا
    0.61
     Aquí
    0.61
    heres
    0.61
    Act Density 0.040%

    No Known Activations