© 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. 10-GEMMASCOPE-RES-131K
    7. 13315
    Prev
    Next
    INDEX
    Explanations

    words after a comma

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

    The neuron fires on sentence‐ or question‐initial tokens (e.g. “Now,” “I,” “How,” “Did”) – in other words, it marks the beginnings of new sentences or questions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     inesperado
    -1.97
    他們
    -1.79
     savage
    -1.72
    になった
    -1.66
    šče
    -1.59
     tangled
    -1.59
    <bos>
    -1.58
    찡
    -1.55
     быстро
    -1.55
     festen
    -1.54
    POSITIVE LOGITS
     in
    2.22
    瘊
    1.66
     beyond
    1.60
     about
    1.56
     how
    1.53
     unless
    1.51
     except
    1.50
    ()));
    1.48
     whether
    1.47
    桖
    1.47
    Activations Density 0.036%

    No Known Activations