© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 119450
    Prev
    Next
    INDEX
    Explanations

    closing code tags

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

    string literals in code, particularly those ending with semicolons.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    This neuron detects string-literal delimiters (the quotation marks that begin/end string constants in code).

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
     
    -2.67
    <em>
    -2.23
     P
    -2.22
     .
    -2.13
     C
    -2.13
     R
    -2.09
     M
    -2.08
    理石
    -2.05
     E
    -1.95
    le
    -1.93
    POSITIVE LOGITS
    鼉
    2.42
    嫑
    2.22
     商店
    2.19
    </b>
    2.03
    矬
    2.02
    耵
    1.99
    เพื่อ
    1.98
    嘥
    1.95
    糨
    1.94
    鑁
    1.92
    Activations Density 0.011%

    No Known Activations