© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 117147
    Prev
    Next
    INDEX
    Explanations

    scientific notation

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

    The neuron activates on numeric tokens—especially digits and numbers appearing in scientific formulas or 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_34/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
    誹
    -1.02
     although
    -0.99
    諜
    -0.96
    してます
    -0.94
    有難
    -0.90
    ｧ
    -0.90
    してました
    -0.87
     despite
    -0.87
    游戏亮点
    -0.87
     only
    -0.86
    POSITIVE LOGITS
     插画
    0.93
    irée
    0.90
    peper
    0.90
    還
    0.89
     فريبيس
    0.89
     Hacer
    0.89
     ß
    0.88
    tition
    0.88
     quella
    0.87
    pedia
    0.86
    Activations Density 0.029%

    No Known Activations