© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 40-GEMMASCOPE-2-RES-65K
    7. 50128
    Prev
    Next
    INDEX
    Explanations

    user queries and platform expansion

    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_40_width_65k_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
     Conversely
    0.40
    婼
    0.38
    Fatalf
    0.38
    盎
    0.38
    푎
    0.36
    Conversely
    0.36
     Interfer
    0.36
    हेलो
    0.35
     എന്നാൽ
    0.35
    یرے
    0.35
    POSITIVE LOGITS
    文字
    0.41
     λειτουργ
    0.40
     Apps
    0.40
    
    0.38
     assunto
    0.38
     humbly
    0.38
     दूसरा
    0.38
    erende
    0.37
     rook
    0.37
     inequ
    0.37
    Activations Density 0.002%

    No Known Activations