© 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-3-27B-IT
    5. 31-GEMMASCOPE-2-RES-262K
    7. 3983
    Prev
    Next
    INDEX
    Explanations

    after very the of this

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_31_width_262k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    すべき
    0.47
    Grilled
    0.46
    ゼロ
    0.45
     overload
    0.44
    癀
    0.44
    kcal
    0.43
    Exceptionally
    0.43
     இன்
    0.42
    ljanje
    0.42
     overwhel
    0.42
    POSITIVE LOGITS
     L
    0.45
     tế
    0.44
    em
    0.43
     B
    0.43
     con
    0.43
     as
    0.43
    ्राम
    0.43
     V
    0.42
    س
    0.42
     مع
    0.41
    Activations Density 0.000%

    No Known Activations