© 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. 169552
    Prev
    Next
    INDEX
    Explanations

    identifying underlying themes

    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
     என்கிறார்
    0.42
    klass
    0.41
    ylist
    0.40
    령
    0.40
    timetable
    0.40
    ългар
    0.40
    首次
    0.39
     सध्या
    0.39
    kontakt
    0.38
    ≧
    0.38
    POSITIVE LOGITS
     tangible
    0.45
     temporary
    0.42
    этому
    0.40
     optionally
    0.40
     comforting
    0.39
     faithfully
    0.39
    omeness
    0.39
     mitigating
    0.38
    Won
    0.38
     temporarily
    0.38
    Activations Density 0.007%

    No Known Activations