© 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. 169308
    Prev
    Next
    INDEX
    Explanations

    preparing and improving

    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
     Rico
    0.41
    ائض
    0.41
     salir
    0.39
     Millennials
    0.38
     c
    0.38
     Aire
    0.37
     Embr
    0.37
     ilan
    0.37
    anter
    0.37
     OECD
    0.37
    POSITIVE LOGITS
    preparing
    0.46
    Selecting
    0.46
    preparation
    0.44
    指定的
    0.42
    Preparing
    0.42
    昔
    0.42
    憸
    0.41
     Selecting
    0.41
    Prepare
    0.41
    Fraud
    0.41
    Activations Density 0.002%

    No Known Activations