© 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. 215042
    Prev
    Next
    INDEX
    Explanations

    hierarchical attention structure

    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
    ＇
    0.48
    પાસ
    0.46
    łach
    0.44
     تس
    0.43
    俬
    0.41
     secretly
    0.40
     元
    0.40
     पेंसिल
    0.39
     Kode
    0.39
     ఛ
    0.39
    POSITIVE LOGITS
    amingo
    0.39
     attention
    0.38
     atenção
    0.37
    ប្រ
    0.36
    Step
    0.36
     Sanctuary
    0.36
    appearance
    0.35
    height
    0.35
     cares
    0.35
    smaller
    0.34
    Activations Density 0.000%

    No Known Activations