© 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. 128212
    Prev
    Next
    INDEX
    Explanations

    This neuron detects closing or end‐of‐block tokens and their indentation (e.g. XML/HTML closing tags like “</ItemGroup>,” self‐closing tags “/>,” CSS “},” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    web resources and code snippets

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    打量
    -0.96
    ǿ
    -0.91
     courtier
    -0.89
    🇧
    -0.88
    taş
    -0.88
     الأخرى
    -0.87
     ļoti
    -0.86
     philosophic
    -0.85
     這
    -0.84
    wki
    -0.84
    POSITIVE LOGITS
     instead
    0.92
    而是
    0.87
     tables
    0.84
     resources
    0.82
    das
    0.81
    人都
    0.80
    Table
    0.80
     ="";
    0.80
     drawings
    0.79
    ATED
    0.79
    Activations Density 0.002%

    No Known Activations