© 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. 41-GEMMASCOPE-2-RES-262K
    7. 4429
    Prev
    Next
    INDEX
    Explanations

    Japan and Japanese contexts

    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_all/layer_41_width_262k_l0_big
    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
    鎊
    0.76
    िरिक्स
    0.74
    ឡិច
    0.73
    🥯
    0.73
    ROTTLE
    0.73
    軛
    0.72
    َان
    0.71
    颶
    0.71
     फेव
    0.71
    瘓
    0.71
    POSITIVE LOGITS
     Japanese
    2.56
     Japan
    2.53
     일본
    2.38
    Japan
    2.34
    Japanese
    2.27
     Tokyo
    2.23
     япон
    2.14
     Jepang
    2.13
     japan
    2.06
     Osaka
    2.02
    Activations Density 0.097%

    No Known Activations