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

    Neuronpedia

    INDEX
    Explanations

    interactions or inquiries related to assistance and suggestions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    asking a question and expecting an answer

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

    The neuron detects markers that indicate an answer or reply/closing in a post (e.g., "A:", "Thanks", and similar reply/closing tokens).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
     AssemblyTitle
    -0.68
     autorytatywna
    -0.58
    UserScript
    -0.54
    Autoritní
    -0.49
    المناصب
    -0.48
    씩
    -0.48
    <pad>
    -0.48
     firebaseConfig
    -0.47
    <unused20>
    -0.47
    [@BOS@]
    -0.47
    POSITIVE LOGITS
    RenderAtEndOf
    0.38
    GEBURTS
    0.36
     województ
    0.32
    expandindo
    0.32
    antMatchers
    0.32
     ovary
    0.31
     незавершена
    0.31
     везе
    0.31
    complain
    0.31
     compor
    0.30
    Act Density 0.082%

    No Known Activations