© 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. 129377
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on numeric tokens, i.e. sequences of digits.

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

    Congenital, Founded, May, Stud

    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
    AppRoutingModule
    -1.19
    事で
    -1.09
     Our
    -1.06
    DeclareMath
    -1.05
     nasze
    -1.03
    淡淡
    -1.02
    uristic
    -1.01
    久しぶり
    -1.01
     OGSÅ
    -1.00
     serce
    -1.00
    POSITIVE LOGITS
     now
    1.04
     Nitrate
    0.85
     пока
    0.84
     Straw
    0.82
     Heav
    0.81
     except
    0.80
    tagena
    0.79
    ’
    0.79
     Sumatra
    0.79
     in
    0.78
    Activations Density 0.003%

    No Known Activations