© 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. 169297
    Prev
    Next
    INDEX
    Explanations

    code comparisons and results

    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
     remporté
    0.44
     ಮಾಡಲು
    0.43
     』
    0.41
     //=>
    0.39
     살펴보도록
    0.38
     겁니다
    0.37
     {});
    0.37
    ശാ
    0.37
     статус
    0.35
    );
    0.35
    POSITIVE LOGITS
    Ler
    0.41
     spelling
    0.39
     Spelling
    0.37
     Augustus
    0.36
     garr
    0.36
    磊
    0.36
    }},
    0.36
    Fus
    0.36
     hur
    0.35
    Pear
    0.35
    Activations Density 0.016%

    No Known Activations