© 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. 34-GEMMASCOPE-RES-131K
    7. 77127
    Prev
    Next
    INDEX
    Explanations

    William, British, Kawasaki, Structure, sizeof_, Cor, An

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

    The neuron activates on numeric tokens, especially sequences of digits (e.g. years, ages, counts).

    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
     ملک
    -0.81
     Erkrank
    -0.74
    大家好
    -0.70
     pidana
    -0.69
     szükség
    -0.69
    upassen
    -0.69
     lưng
    -0.66
    目指
    -0.66
    ponses
    -0.66
     Dominican
    -0.66
    POSITIVE LOGITS
     British
    1.05
     Company
    0.97
    British
    0.95
     mutiny
    0.90
    company
    0.87
     Warren
    0.86
     Mut
    0.85
    شر
    0.85
     british
    0.84
    Company
    0.83
    Activations Density 0.015%

    No Known Activations