© 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. 34516
    Prev
    Next
    INDEX
    Explanations

    controller action and format

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

    The neuron strongly fires on colon characters (":"), especially in code and URL contexts.

    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
    andosi
    -0.85
     במק
    -0.85
    hattan
    -0.81
    geordnete
    -0.78
     requieren
    -0.77
     berdiri
    -0.76
     potom
    -0.76
     contienen
    -0.73
    siyah
    -0.73
    绎
    -0.73
    POSITIVE LOGITS
    format
    1.25
     format
    1.01
    controller
    1.00
    constraints
    0.97
     Format
    0.93
     constraints
    0.93
     formats
    0.92
     xhr
    0.92
    Format
    0.91
     id
    0.88
    Activations Density 0.026%

    No Known Activations