© 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. 24-GEMMASCOPE-2-TRANSCODER-262K
    7. 176207
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_24_width_262k_l0_small_affine
    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
     narrates
    0.53
     coorden
    0.52
    氧
    0.52
     istraž
    0.51
     izme
    0.50
    編輯
    0.50
     encontraba
    0.50
     gurus
    0.49
     explores
    0.49
     intimidate
    0.49
    POSITIVE LOGITS
    o
    0.61
    or
    0.54
    Y
    0.53
    in
    0.52
    ir
    0.50
    e
    0.50
    er
    0.49
    pu
    0.49
    ften
    0.49
    film
    0.49
    Activations Density 0.000%

    No Known Activations