© 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. 22-GEMMASCOPE-RES-131K
    7. 15131
    Prev
    Next
    INDEX
    Explanations

    date ranges, start, end

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

    The neuron fires on numeric literals (pure digit tokens) 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
     خونه
    -1.00
    ↵
    -0.97
     incontri
    -0.96
     Jokowi
    -0.96
     ways
    -0.93
    鋅
    -0.93
     spus
    -0.92
     choices
    -0.91
    忘れ
    -0.90
     reprezent
    -0.88
    POSITIVE LOGITS
     period
    1.68
     duration
    1.63
     startDate
    1.34
     months
    1.34
    startDate
    1.24
     date
    1.20
     time
    1.18
    ťaž
    1.16
     dates
    1.15
    ciled
    1.15
    Activations Density 0.012%

    No Known Activations