© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 104716
    Prev
    Next
    INDEX
    Explanations

    code delimiters or block starts

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

    quotation marks or forward slashes in programming or technical code contexts.

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

    The neuron fires on tokens that combine a quotation mark with another punctuation character (e.g. '"/', '/"', '|', etc.), effectively spotting string‐delimiter fragments.

    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
    We
    -2.52
     Preferably
    -2.19
    .
    -2.14
    的確
    -2.11
     But
    -2.05
    這是
    -2.03
     sämtliche
    -2.02
     грудня
    -2.00
    नलोड
    -1.99
    This
    -1.95
    POSITIVE LOGITS
    他自己
    2.34
     эту
    2.16
     they
    2.08
    教你
    2.08
     viaja
    2.06
     técnico
    2.05
     blockbuster
    2.05
     gründ
    2.03
    ỽ
    1.99
     poética
    1.97
    Activations Density 0.019%

    No Known Activations