© 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. 58952
    Prev
    Next
    INDEX
    Explanations

    scene builder or scene graph

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

    The neuron strongly responds to occurrences of the word “scene” (e.g. in JavaFX import or class names).

    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
     with
    -0.96
     blatantly
    -0.94
     brutal
    -0.92
     intricacies
    -0.91
     during
    -0.89
     taking
    -0.88
     using
    -0.86
     unrelenting
    -0.85
     reimburse
    -0.85
     incorporate
    -0.85
    POSITIVE LOGITS
     rè
    1.02
     magnificent
    0.95
     tremendous
    0.95
     VERY
    0.93
    whose
    0.92
    control
    0.92
     breathtaking
    0.92
     ww
    0.91
    subscribed
    0.90
    EFFECT
    0.89
    Activations Density 0.002%

    No Known Activations