© 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. 10-GEMMASCOPE-RES-131K
    7. 84002
    Prev
    Next
    INDEX
    Explanations

    changes with

    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_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
     even
    -1.51
    even
    -1.39
     to
    -1.37
     if
    -1.30
     what
    -1.22
     for
    -1.19
     confirma
    -1.17
     but
    -1.16
     didn
    -1.15
     même
    -1.11
    POSITIVE LOGITS
     huden
    1.45
     tänka
    1.38
     förstå
    1.38
     verkligen
    1.36
     föres
    1.34
     särsk
    1.32
     vilja
    1.31
     barnen
    1.28
     möjlighet
    1.28
     parfü
    1.28
    Activations Density 0.069%

    No Known Activations