© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 95409
    Prev
    Next
    INDEX
    Explanations

    HTML tags structure

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

    The neuron selectively activates on tokens that include numeric digits (e.g. version numbers, numeric literals, or any token containing “0–9”).

    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
     começou
    -0.99
     privacidad
    -0.97
    ugg
    -0.96
     abriu
    -0.93
     origem
    -0.93
     execução
    -0.93
     atendimento
    -0.92
     Tecnología
    -0.92
     pulseira
    -0.92
    chaine
    -0.92
    POSITIVE LOGITS
    このような
    1.17
     playful
    0.99
     blissful
    0.96
     decreasing
    0.92
    intera
    0.91
     शुर
    0.89
    BLO
    0.88
     gela
    0.87
    そのような
    0.87
    arParams
    0.87
    Activations Density 0.004%

    No Known Activations