© 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. 33437
    Prev
    Next
    INDEX
    Explanations

    percentage, census, Q codes

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

    The neuron responds most strongly to numeric data—especially percentages (numbers in parentheses followed by “%”)—in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     on
    -1.48
     their
    -1.45
     just
    -1.35
     will
    -1.32
     (
    -1.31
     trying
    -1.31
     with
    -1.27
     even
    -1.26
     all
    -1.23
    (
    -1.17
    POSITIVE LOGITS
     سات
    1.28
    푉
    1.28
     다른
    1.26
    Etape
    1.23
    Adverten
    1.22
    CurtirCurtir
    1.20
    GDPR
    1.20
    sprzedam
    1.20
    ぽん
    1.20
     ส์
    1.19
    Activations Density 0.000%

    No Known Activations