© 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. 129213
    Prev
    Next
    INDEX
    Explanations

    The neuron detects numerical tokens (e.g. page, volume, year, and reference numbers) in patent‐style citations and technical listings.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    page numbers and citations

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     completamente
    -1.08
     fonctionner
    -1.02
     lampa
    -0.94
    󠁬
    -0.93
    Beginning
    -0.93
    rire
    -0.92
     permanecer
    -0.92
     boden
    -0.91
     comporta
    -0.90
     jTextField
    -0.90
    POSITIVE LOGITS
    그러나
    0.98
    镭
    0.95
     शुर
    0.88
     vieja
    0.88
     asumi
    0.85
     both
    0.83
     contributes
    0.81
    قیمت
    0.81
    oscopic
    0.80
     stasiun
    0.79
    Activations Density 0.001%

    No Known Activations