© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 854
    Prev
    Next
    INDEX
    Explanations

    rectangle, Rect, rect, rectangles

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

    The neuron detects code tokens related to view/layout geometry calls and initializer methods (e.g. CGRect, initWithFrame:, setFrame:, drawRect:).

    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
    払
    -0.71
     Graduation
    -0.70
    removal
    -0.67
    बाट
    -0.67
    Rigidbody
    -0.66
    Guyana
    -0.66
    Graduation
    -0.65
    ენ
    -0.64
    localScale
    -0.64
    kkel
    -0.63
    POSITIVE LOGITS
     rectangular
    3.86
     rectangle
    3.55
    Rect
    3.53
     rect
    3.48
     Rectangle
    3.36
     Rect
    3.28
     rectangles
    3.19
     Rectangular
    3.14
    rect
    3.08
    Rectangle
    3.08
    Activations Density 0.074%

    No Known Activations