© 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. 40239
    Prev
    Next
    INDEX
    Explanations

    appendix B

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

    The neuron activates on numeric tokens (especially decimal or integer values such as section numbers, measurements, or data counts).

    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
     implique
    -1.23
     เปิด
    -1.13
    ทีม
    -1.04
    เริ่ม
    -1.03
     casado
    -1.02
     陈
    -1.01
    fordable
    -1.00
     bantuan
    -0.99
     couverture
    -0.98
    ыр
    -0.96
    POSITIVE LOGITS
     list
    1.29
     sample
    1.23
    onymus
    1.16
     }}"></
    1.16
    alisis
    1.09
     to
    1.09
     writing
    1.05
     additional
    1.04
    álise
    1.04
    erobic
    1.01
    Activations Density 0.003%

    No Known Activations