© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 40318
    Prev
    Next
    INDEX
    Explanations

    code structure delimiters

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

    The neuron strongly activates on identifier tokens that are part of package/module import or namespace declarations (e.g. library names or path components in “import”/“package” statements).

    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
    ynthetic
    -2.17
    های
    -2.11
    ſelves
    -2.08
    洐
    -2.08
    縐
    -2.06
     slated
    -2.05
    hooter
    -2.02
     dévelo
    -2.02
    两年
    -1.99
    妗
    -1.96
    POSITIVE LOGITS
     erstes
    2.45
    pecabe
    2.27
    嘡
    2.23
    IntentFilter
    2.23
    <bos>
    2.19
    铪
    2.17
     kleines
    2.16
     übri
    2.16
    (\
    2.16
     sogen
    2.14
    Activations Density 0.245%

    No Known Activations