© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-27B-IT
    3. 31-GEMMASCOPE-2-RES-262K
    4. 30653
    Prev
    Next
    INDEX
    Explanations

    list, number at the beginning/end of the list.duplicates/weak points and selecting the top 5 featuresstatement that *must* be trueMinimized antibiotic exposure. * Limitations: * Not for PyelonephritisRequires more setup and/or specific equipmentRecovered well after a less-than-stellar startday to night gives them the edgeskirt, specifically a pleated, micro, and miniskirtStyles I can *adopt* based on your prompting

    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
    摶
    0.42
    まとめ
    0.41
    အတွက်
    0.39
     包括
    0.39
    まとめて
    0.38
     postponing
    0.37
     VE
    0.37
     बेकार
    0.37
     _{\
    0.37
     формирование
    0.36
    POSITIVE LOGITS
     videa
    0.38
    هما
    0.38
    uki
    0.37
    تن
    0.37
    したり
    0.36
    ‍या
    0.36
    irrah
    0.35
    이나
    0.34
    vora
    0.33
    罅
    0.33
    Activations Density 0.019%

    No Known Activations