© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128811
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively fires on standalone numeric tokens (whole numbers), e.g. the counts or scores at the ends of review/testimonial entries.

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

    thank you / read more

    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
     (
    -1.78
     with
    -1.58
     in
    -1.42
     of
    -1.23
     for
    -1.22
     because
    -1.22
     into
    -1.19
     on
    -1.16
     to
    -1.16
    -
    -1.16
    POSITIVE LOGITS
    ⽼
    1.72
    Ꮠ
    1.38
    شكرا
    1.38
    ㌧
    1.34
    鵙
    1.34
    킁
    1.34
    綺麗な
    1.33
     起毛
    1.33
     vacunas
    1.32
     (!__
    1.31
    Activations Density 0.004%

    No Known Activations