© 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. 31-GEMMASCOPE-2-RES-262K
    7. 137376
    Prev
    Next
    INDEX
    Explanations

    historical puzzle, philosophical traditions

    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_31_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
     optimized
    0.43
     сма
    0.43
     optimization
    0.42
    optimization
    0.42
    regation
    0.41
    toolStrip
    0.41
    酰
    0.40
     precipitation
    0.40
    최
    0.40
    습니다
    0.39
    POSITIVE LOGITS
     bygone
    0.53
     quieren
    0.48
     muốn
    0.47
     strolled
    0.46
     fortunes
    0.46
     fashionable
    0.46
    検索
    0.46
     nuov
    0.46
     ګرځنده
    0.46
     longed
    0.44
    Activations Density 0.008%

    No Known Activations