© 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. 128717
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on multi‐digit numeric tokens (i.e. numbers).

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

    title or start of descriptive phrase

    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
     acara
    -0.95
     filós
    -0.92
     Corintios
    -0.91
     saca
    -0.90
    ઃ
    -0.90
     étanche
    -0.89
     {
    -0.89
    禁止
    -0.85
    пля
    -0.84
     вроде
    -0.83
    POSITIVE LOGITS
    öglichkeiten
    1.01
    Pontos
    0.97
    もともと
    0.95
    Recept
    0.95
    Mails
    0.93
    元々
    0.93
    Як
    0.93
    Kontak
    0.90
    Antwoord
    0.89
     放題
    0.89
    Activations Density 0.000%

    No Known Activations