© 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. 68040
    Prev
    Next
    INDEX
    Explanations

    column and numbers

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

    The neuron fires on numeric tokens (multi-digit numbers), especially values like pixel coordinates, sizes, delays, or other quantitative parameters.

    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
     çanta
    -1.13
     koltuk
    -1.13
     yeşil
    -1.09
    Already
    -1.06
     taşlı
    -1.05
    mathrm
    -0.98
    تج
    -0.97
     konfigura
    -0.97
     myn
    -0.96
    موقع
    -0.96
    POSITIVE LOGITS
     (
    1.02
     CITED
    0.91
     still
    0.91
    ですが
    0.89
    色が
    0.89
     Under
    0.86
    虜
    0.84
     Just
    0.84
    幹線
    0.83
     буквально
    0.83
    Activations Density 0.001%

    No Known Activations