© 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. 58539
    Prev
    Next
    INDEX
    Explanations

    DHS reports

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

    The neuron specializes in picking out numerical or statistical references (digits, measurements, counts, percentages) 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_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
     raso
    -0.75
     но
    -0.71
     Mike
    -0.71
     ellos
    -0.70
     обли
    -0.70
    städ
    -0.69
     Americans
    -0.69
     warmen
    -0.69
    atoga
    -0.69
    ismas
    -0.67
    POSITIVE LOGITS
     DHS
    0.92
     fecund
    0.86
     biomarker
    0.85
    avimas
    0.83
     Bina
    0.82
     oth
    0.81
     rectangles
    0.80
     whar
    0.78
     omnis
    0.77
     assoc
    0.77
    Activations Density 0.043%

    No Known Activations