© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 38480
    Prev
    Next
    INDEX
    Explanations

    Berlin Wall, Phasmida, CNHS, TT8, vastator, NEXBOX, Jao, Full text, Picone, Jiroenbashi, FlatOut, Black Pond

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron activates on uppercase letter sequences—especially acronyms, abbreviations, and other all‐caps or capitalized tokens.

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     keď
    -1.46
    ᨆ
    -1.41
    暖かい
    -1.35
    桫
    -1.33
     dolay
    -1.31
    鴴
    -1.28
    ふわふわ
    -1.27
    鳚
    -1.25
    ふわ
    -1.24
     grandiose
    -1.23
    POSITIVE LOGITS
     που
    1.45
     của
    1.43
     have
    1.42
     decoraciones
    1.34
    ();
    1.34
    ที่
    1.34
    1.28
    _{\
    1.27
     are
    1.23
     other
    1.23
    Activations Density 0.047%

    No Known Activations