© 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. 10-GEMMASCOPE-RES-131K
    7. 10929
    Prev
    Next
    INDEX
    Explanations

    language, -ism, curvature

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

    This neuron responds to long, specialized terms—especially scientific or technical words (including discipline names, diseases, and other formal vocabulary).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     which
    -2.61
     still
    -2.00
     allows
    -1.74
     ensures
    -1.71
     allow
    -1.70
     mientras
    -1.70
     seems
    -1.64
     enables
    -1.60
     Some
    -1.59
     now
    -1.59
    POSITIVE LOGITS
    
    1.62
     ongel
    1.51
    РАВ
    1.50
    鲧
    1.50
    РЯ
    1.50
     墊
    1.49
    啐
    1.45
     唯美
    1.41
    ܆
    1.41
    könyv
    1.34
    Activations Density 0.226%

    No Known Activations