© 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. 40-GEMMASCOPE-2-RES-262K
    7. 214960
    Prev
    Next
    INDEX
    Explanations

    security and privacy concerns

    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_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
     repres
    0.42
     evanes
    0.41
    दर्शिता
    0.41
    쿄
    0.41
     emir
    0.39
     forne
    0.39
     көр
    0.38
     Tem
    0.38
    emin
    0.38
    펌
    0.38
    POSITIVE LOGITS
    security
    0.60
    SECURITY
    0.59
     security
    0.55
    Security
    0.55
     SECURITY
    0.52
    セキュリティ
    0.50
     keamanan
    0.49
     Security
    0.48
     सुरक्षा
    0.46
    誕生日
    0.45
    Activations Density 0.003%

    No Known Activations