© 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. 22-GEMMASCOPE-RES-131K
    4. 36293
    Prev
    Next
    INDEX
    Explanations

    common phrases or code keywords

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

    The neuron fires on programming‐language syntax tokens—especially reserved words (like register, readonly, define) and punctuation (notably the “(” token) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     already
    -1.06
     Already
    -1.05
     what
    -1.05
    Already
    -0.99
     it
    -0.99
     everything
    -0.96
     уже
    -0.95
     its
    -0.91
     Bereits
    -0.89
    baris
    -0.88
    POSITIVE LOGITS
    économ
    1.03
    等は
    1.00
    ffold
    1.00
    excellents
    0.98
    siamo
    0.96
    بندی
    0.95
     pitié
    0.93
    Hilsen
    0.92
     появится
    0.91
    TEACH
    0.91
    Activations Density 0.000%

    No Known Activations