© 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. 32854
    Prev
    Next
    INDEX
    Explanations

    version 2, arrested state, bind to

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

    XML, code documentation, and markup syntax in programming files.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron strongly responds to C-style code tokens—especially C# XML-doc “param” tags, Boolean literal “false,” and method signature keywords like “protected” and “void.”

    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
     PlayStation
    -1.02
     volta
    -0.87
    EditorGUILayout
    -0.78
    änd
    -0.77
     Leonardo
    -0.77
     Romance
    -0.77
     ciga
    -0.76
    \[
    -0.75
     dữ
    -0.75
     Bourbon
    -0.75
    POSITIVE LOGITS
    ilish
    0.80
     zuges
    0.79
     klä
    0.75
    に住
    0.73
    perfil
    0.73
    LLABLE
    0.72
    tka
    0.71
    elesaikan
    0.69
     proveedor
    0.68
     พร้อม
    0.68
    Activations Density 0.001%

    No Known Activations