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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 81717
    Prev
    Next
    INDEX
    Explanations

    expressions of thanks and personal words

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

    This neuron responds strongly to punctuation marks (especially sentence-ending ones like periods, exclamation points, and ellipses) and other high-frequency tokens that mark emphasis or sentence boundaries.

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
     at
    -0.93
     egli
    -0.87
    为什么
    -0.87
     with
    -0.87
     этим
    -0.85
     iPad
    -0.85
     почему
    -0.85
     Why
    -0.82
    まだ
    -0.82
    іє
    -0.82
    POSITIVE LOGITS
     words
    2.06
     thank
    1.99
     your
    1.73
    Without
    1.71
    Words
    1.70
     Благодаря
    1.70
     Thank
    1.62
     without
    1.60
     THANK
    1.53
     Words
    1.48
    Activations Density 0.021%

    No Known Activations