© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 104672
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions involving numbers and exponents

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

    This neuron responds primarily to numeric tokens (numbers and digits) in the 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_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
     poë
    -1.13
    ά
    -1.09
    Tembelea
    -1.08
     kado
    -1.07
     lumineux
    -1.04
    ඈ
    -1.04
     péro
    -1.03
     réserver
    -1.03
     portas
    -1.03
    往前
    -1.02
    POSITIVE LOGITS
     at
    1.02
     had
    0.95
     around
    0.93
     ontstaan
    0.93
     alongside
    0.93
     just
    0.92
     when
    0.91
     cleverly
    0.89
    我们就
    0.89
     blijkt
    0.87
    Activations Density 0.147%

    No Known Activations