© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 68108
    Prev
    Next
    INDEX
    Explanations

    specific beginnings of phrases

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

    The neuron selectively activates on numeric tokens (years, page/volume numbers, digits).

    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
    angal
    -0.90
    gion
    -0.88
    epsfig
    -0.87
     ͡°)
    -0.85
    زشکی
    -0.85
     gevo
    -0.83
    疲れ
    -0.83
     Steak
    -0.82
    的名字
    -0.82
    ائر
    -0.82
    POSITIVE LOGITS
    tôi
    0.91
    رحلة
    0.87
    Chỉ
    0.86
    khó
    0.85
    IFT
    0.85
    Nếu
    0.84
    OPA
    0.84
    توضیحات
    0.84
    電腦
    0.83
     неожидан
    0.83
    Activations Density 0.000%

    No Known Activations