© 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. 168258
    Prev
    Next
    INDEX
    Explanations

    cost per person

    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
     हाद
    0.44
     बढ़ाता
    0.43
    -{\
    0.42
    每年
    0.41
     цена
    0.40
    Annual
    0.39
     annually
    0.39
     jähr
    0.39
     цену
    0.39
    別途
    0.38
    POSITIVE LOGITS
    两个人
    0.66
    二人
    0.58
    两人
    0.52
    兩人
    0.49
     दोघा
    0.49
     двох
    0.47
    一个人
    0.46
     둘
    0.46
     two
    0.45
     দুজনের
    0.45
    Activations Density 0.001%

    No Known Activations