© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 26480
    Prev
    Next
    INDEX
    Explanations

    greater than symbol

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

    This neuron fires on markup and code‐annotation tokens—especially punctuation like colons in reST field lists and backslash-escaped commands or numeric references in technical documentation.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    <bos>
    -2.31
     with
    -1.98
     where
    -1.96
     at
    -1.87
     because
    -1.83
     in
    -1.72
     when
    -1.71
    ちなみに
    -1.66
     which
    -1.66
    を
    -1.63
    POSITIVE LOGITS
    為に
    2.03
     embaraz
    1.77
    튿
    1.73
     autén
    1.69
    ところに
    1.67
    𓆏
    1.67
    に入れる
    1.66
    たびに
    1.66
    すっ
    1.66
     sám
    1.64
    Activations Density 0.035%

    No Known Activations