© 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. 10-GEMMASCOPE-RES-131K
    7. 112744
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions and equations

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

    This neuron activates on numeric literals (especially decimal numbers) in code or math 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_10/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
    carnated
    -1.37
     such
    -1.30
    biendo
    -1.23
    coglie
    -1.23
    躄
    -1.18
     but
    -1.18
     ThemeData
    -1.17
    рных
    -1.16
    ListTile
    -1.14
    cagon
    -1.14
    POSITIVE LOGITS
     anvä
    1.26
     fratelli
    1.21
     Gefühl
    1.21
     とき
    1.19
    huevos
    1.16
     igreja
    1.16
    Them
    1.16
     Kenntnis
    1.15
    zapatillas
    1.13
    dibujos
    1.13
    Activations Density 0.172%

    No Known Activations