© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 26864
    Prev
    Next
    INDEX
    Explanations

    math and physics variables

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

    This neuron activates on numeric tokens (especially numbers in equations or as subscripts/exponents).

    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
     noël
    -1.08
     sweter
    -1.02
     eV
    -1.00
     bienven
    -0.97
    Uhr
    -0.96
     jeter
    -0.94
     hadas
    -0.93
     dîner
    -0.93
     bús
    -0.93
    州市
    -0.93
    POSITIVE LOGITS
    I
    1.01
    での
    0.98
    u
    0.96
     tip
    0.93
    L
    0.92
    with
    0.90
    chism
    0.90
     especially
    0.90
     No
    0.88
     first
    0.88
    Activations Density 0.007%

    No Known Activations