© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26
    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-27B-IT
    3. 11-GEMMASCOPE-2-TRANSCODER-262K
    4. 34994
    Prev
    Next
    INDEX
    Explanations

    rolls in, a greasy shroudstorm rolled in. The wavesit rolled in – a thick, oilyThe light didn’t sweep in a predictable arc. Sometimes it was bright, sometimes dim, sometimes it seemed to be in *multiple* places at oncea new tide is rolling in, and it’s made of siliconseems to press in on all sides. The buildings are old, mostly clapboard and stone, weathered to a uniform, unsettling pallorThe wind came in strong, blowing the curtains back. It was a cool air, a bit of a chill, but somehow it felt good

    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_11_width_262k_l0_small_affine
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
    sset
    0.70
    hér
    0.63
     coval
    0.62
    íduos
    0.61
    s
    0.61
    o
    0.60
     bialgebras
    0.58
    ም
    0.57
     районы
    0.57
    iť
    0.55
    POSITIVE LOGITS
    umbling
    0.64
    usk
    0.63
     t
    0.61
    ua
    0.58
    ulu
    0.55
     O
    0.55
     o
    0.54
     by
    0.54
    arsk
    0.54
     or
    0.53
    Activations Density 0.011%

    No Known Activations