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

    Neuronpedia

    INDEX
    Explanations

    scientific publications

    np_max-act · gemini-2.0-flash

    technical scientific/medical terms, especially those with Latin/Greek morphological patterns common in biomedical and taxonomic language

    oai_token-act-pair · gpt-5Triggered by @chenshw0109

    the neuron detects technical scientific/medical vocabulary (specialized biological, medical, or technical terms).

    oai_token-act-pair · gpt-5-miniTriggered by @chenshw0109
    New Auto-Interp
    Negative Logits
    unk
    -0.60
    inov
    -0.57
    en
    -0.56
     Plu
    -0.54
    ण
    -0.53
    ation
    -0.46
    <eos>
    -0.46
     information
    -0.46
     seyn
    -0.45
    ing
    -0.44
    POSITIVE LOGITS
     CreateTagHelper
    0.94
    sonaro
    0.84
    úgó
    0.79
     EconPapers
    0.78
     يتيمه
    0.77
     חיצוניים
    0.76
    batore
    0.75
    antMatchers
    0.73
    Hauptartikel
    0.73
     estekak
    0.72
    Act Density 0.231%

    No Known Activations