© 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-262K
    7. 11078
    Prev
    Next
    INDEX
    Explanations

    descriptive qualities

    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
    ،
    0.65
    ，
    0.59
    ﹐
    0.50
     ،
    0.49
     ,
    0.48
    、
    0.48
    ፣
    0.46
     considéré
    0.43
     த
    0.42
    ؛
    0.42
    POSITIVE LOGITS
    but
    0.60
     but
    0.56
     მაგრამ
    0.53
     लेकिन
    0.51
    non
    0.50
     mutta
    0.49
    multi
    0.49
     multimillion
    0.47
     nhưng
    0.47
     albeit
    0.47
    Activations Density 0.024%

    No Known Activations