© 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. 10-GEMMASCOPE-RES-131K
    7. 128480
    Prev
    Next
    INDEX
    Explanations

    code identifiers and punctuation

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

    This neuron responds to code‐related punctuation and operator tokens (e.g. colons, dots, double colons, parentheses, attribute quotes), i.e. non-alphanumeric syntax symbols.

    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
     arbeta
    -1.70
    攵
    -1.62
    ๘
    -1.60
    饜
    -1.56
     harika
    -1.56
     dekoratif
    -1.55
     動画
    -1.55
     accompagner
    -1.55
    鼉
    -1.52
     jouw
    -1.51
    POSITIVE LOGITS
     It
    1.77
     so
    1.75
     Although
    1.73
     by
    1.70
    ()}>
    1.64
     There
    1.58
    providedIn
    1.55
     Despite
    1.54
     These
    1.52
     Nevertheless
    1.52
    Activations Density 0.063%

    No Known Activations