© 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. 22-GEMMASCOPE-RES-131K
    7. 75224
    Prev
    Next
    INDEX
    Explanations

    line plot, blue polygon, historical comparator

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

    The neuron strongly activates on mentions of graphical plot elements (dots, lines, polygons) often coupled with color names (e.g. red, blue, green).

    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
     السياس
    -1.06
    からお
    -1.00
     konusunda
    -0.98
    (!__
    -0.98
    ことができる
    -0.98
     assic
    -0.96
    crivez
    -0.95
     caratter
    -0.95
     streets
    -0.94
    ในการ
    -0.93
    POSITIVE LOGITS
     for
    1.43
     without
    1.14
     from
    1.11
    aucune
    1.05
     on
    1.05
    Vegeu
    1.01
     '/':
    0.98
     sypial
    0.97
     contorno
    0.97
     udang
    0.95
    Activations Density 0.008%

    No Known Activations