© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 127775
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly responds to numeric tokens or quantitative expressions (digits, measurements, counts, etc.).

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408

    spoke, patients, run, stronger, active, earth

    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
    Negative Logits
     and
    -1.64
    為に
    -1.42
    明日は
    -1.23
    で見る
    -1.22
    だけではなく
    -1.20
    事で
    -1.20
    必要がある
    -1.19
    すっ
    -1.19
    際には
    -1.19
    さぁ
    -1.16
    POSITIVE LOGITS
     because
    1.17
    だが
    1.15
    參考資料
    1.13
    觑
    1.13
     when
    1.09
    .
    1.09
     if
    1.09
    fourths
    1.08
    も
    1.08
    ebly
    1.07
    Activations Density 0.124%

    No Known Activations