© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 64586
    Prev
    Next
    INDEX
    Explanations

    code blocks with parentheses and braces

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

    The neuron activates primarily on numeric literals (numbers) 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
    пр
    -0.94
    icate
    -0.94
     бак
    -0.93
    mitter
    -0.92
    uted
    -0.91
     by
    -0.90
     enseñanza
    -0.90
    TIPOS
    -0.88
    grading
    -0.85
    IActionResult
    -0.85
    POSITIVE LOGITS
     टीम
    0.94
     spoloč
    0.93
     україн
    0.92
     complementar
    0.91
     outra
    0.91
     imaginar
    0.90
    為に
    0.89
     двигатель
    0.89
    Secondly
    0.88
     مجموع
    0.87
    Activations Density 0.001%

    No Known Activations