© 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. 131005
    Prev
    Next
    INDEX
    Explanations

    context and boundaries

    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
     అందు
    0.40
     ءِ
    0.39
    constexpr
    0.37
     ABSORBED
    0.37
     adenine
    0.36
    ாரு
    0.36
    襬
    0.36
    纪录
    0.36
    лити
    0.36
    放棄
    0.36
    POSITIVE LOGITS
     Husband
    0.40
    rique
    0.37
    bicycle
    0.37
    蜘
    0.37
     relation
    0.37
    Cab
    0.37
     밖에
    0.36
     spa
    0.36
     husband
    0.36
     erect
    0.36
    Activations Density 0.003%

    No Known Activations