© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 130337
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on isolated punctuation tokens—especially sentence-ending marks like periods, exclamation points, and question marks.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    sentence ending punctuation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    🍕
    -1.43
    🤨
    -1.41
    🍺
    -1.40
     Belakang
    -1.39
    wwwwwwww
    -1.38
     toit
    -1.37
    🤙
    -1.37
    –
    -1.35
    ❣
    -1.34
    kül
    -1.33
    POSITIVE LOGITS
    !
    1.60
     see
    1.42
     I
    1.39
     wasn
    1.35
     soluções
    1.29
     never
    1.25
    ?
    1.21
    '
    1.20
    再来
    1.20
     either
    1.20
    Activations Density 0.024%

    No Known Activations