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

    Neuronpedia

    INDEX
    Explanations

    specific punctuation marks or symbols commonly used in formatting text or code

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

    PSA Peugeot Citroen

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

    signals that a section heading or paragraph-level label (e.g., a titled or colon-ended section start) is beginning.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    قایناقلار
    -0.63
    expandindo
    -0.59
    msgTypes
    -0.56
     snippetHide
    -0.55
    fromnode
    -0.54
     defaultstate
    -0.54
    ſelves
    -0.53
    OGND
    -0.53
     shayari
    -0.52
    adeloupe
    -0.51
    POSITIVE LOGITS
    The
    0.54
     The
    0.54
     the
    0.52
     a
    0.42
    ">)</
    0.41
    ScopeManager
    0.40
    utilisons
    0.38
    >)</
    0.37
    >=</
    0.36
     كومونز
    0.36
    Act Density 0.005%

    No Known Activations