© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 56933
    Prev
    Next
    INDEX
    Explanations

    section title openers or metadata tags

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

    The neuron is keyed to tokens that occur in document‐level metadata or headings (e.g. HTML “meta” tags, section titles like “early” in a paper title, and Markdown/footnote markers like “*”).

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     Asociación
    -0.88
    ⛺
    -0.86
    Expo
    -0.85
    auction
    -0.84
     televiz
    -0.82
    छ
    -0.81
    askell
    -0.79
     kuat
    -0.79
     оно
    -0.77
     {!
    -0.77
    POSITIVE LOGITS
    form
    0.91
    Ширина
    0.85
     rather
    0.84
    ner
    0.84
     transform
    0.78
     выпуска
    0.77
    ty
    0.77
    而不是
    0.77
    rot
    0.76
    nade
    0.76
    Activations Density 0.000%

    No Known Activations