© 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. 33945
    Prev
    Next
    INDEX
    Explanations

    data quality and accuracy

    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
    힛
    0.39
     exponer
    0.37
     தங்க
    0.36
    stir
    0.36
    produce
    0.36
    持有
    0.35
     exposing
    0.35
     intensifying
    0.35
     intensified
    0.35
    㧋
    0.35
    POSITIVE LOGITS
     accuracy
    1.28
     Accuracy
    1.17
     inaccuracies
    1.14
    accuracy
    1.13
     inaccurate
    1.13
     inaccuracy
    1.11
    Accuracy
    1.07
     errors
    1.05
     completeness
    1.02
     quality
    0.98
    Activations Density 0.021%

    No Known Activations