© 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. 29-GEMMASCOPE-2-TRANSCODER-262K
    7. 182428
    Prev
    Next
    INDEX
    Explanations

    feels like stepping

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_29_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
     Birch
    0.41
    Д
    0.38
    В
    0.38
    》
    0.38
    व
    0.37
     Jorn
    0.37
    Г
    0.37
    »
    0.36
    곤
    0.36
    乐
    0.35
    POSITIVE LOGITS
     cuánto
    0.42
    wek
    0.41
    koľ
    0.41
    ohu
    0.41
    ਟੀ
    0.40
    hrmacht
    0.40
    бенок
    0.39
     आगाज
    0.39
    swagen
    0.38
    quanto
    0.38
    Activations Density 0.003%

    No Known Activations