© 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. 34-GEMMASCOPE-RES-131K
    7. 87000
    Prev
    Next
    INDEX
    Explanations

    code syntax and object-oriented methods

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

    The neuron fires on structural or formatting tokens that mark boundaries in code or markup—e.g. front-matter “---”, braces, semicolons, XML tags like “para”, code separators, and version numbers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    butik
    -0.88
    إذا
    -0.86
    ôtel
    -0.86
    ৩
    -0.81
    TOXIC
    -0.81
    owała
    -0.79
    şu
    -0.79
     blå
    -0.78
    tuin
    -0.78
    |'.'
    -0.77
    POSITIVE LOGITS
     toString
    1.17
    toString
    1.02
     to
    0.92
     ToString
    0.84
     Override
    0.77
    jo
    0.75
    tisation
    0.75
     stratos
    0.75
     +
    0.74
     caráter
    0.73
    Activations Density 0.001%

    No Known Activations