© 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. 12694
    Prev
    Next
    INDEX
    Explanations

    code, questions, and political/commercial terms

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

    The neuron detects tokens enclosed in parentheses—particularly the argument names or values inside single‐argument parentheses.

    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
     solicita
    -1.15
     options
    -1.15
     then
    -1.12
    Options
    -1.03
     настройки
    -1.02
     Options
    -1.01
     Problemen
    -1.01
     everything
    -1.00
    vereign
    -0.99
     estampa
    -0.99
    POSITIVE LOGITS
    lidean
    1.02
    Análisis
    1.02
     fadeIn
    1.01
     schuhe
    1.00
    BLOOD
    0.97
    ilson
    0.95
    нуться
    0.95
     nuages
    0.94
    Encu
    0.93
     gewann
    0.93
    Activations Density 0.001%

    No Known Activations