© 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. 22-GEMMASCOPE-RES-131K
    7. 71105
    Prev
    Next
    INDEX
    Explanations

    structured data or titles

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

    The neuron primarily activates on numeric literals (digits and number tokens) in the text.

    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
     Nedir
    -1.09
    neden
    -1.07
     inspirasi
    -1.06
     diseñadores
    -1.05
     matriz
    -1.05
     אוי
    -1.03
     from
    -1.00
     berät
    -1.00
    “
    -0.98
    glDelete
    -0.98
    POSITIVE LOGITS
     follow
    0.96
     teils
    0.94
    みたいに
    0.93
     incontra
    0.91
     verheir
    0.91
    カラフル
    0.90
     acompan
    0.90
    ダイヤモンド
    0.90
    这么说
    0.89
     pittura
    0.88
    Activations Density 0.011%

    No Known Activations