© 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. 84424
    Prev
    Next
    INDEX
    Explanations

    F HO Z 3 d 0 1

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

    The neuron fires on numeric tokens—digits, years, times, measurements, etc.—i.e. it detects numbers 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
     what
    -1.21
     dobře
    -0.98
    )}$.
    -0.96
     one
    -0.96
    しておく
    -0.95
     erreur
    -0.95
     börjar
    -0.94
     étape
    -0.94
     zufolge
    -0.91
     usługi
    -0.90
    POSITIVE LOGITS
    กล้
    1.10
     ral
    1.09
    Leírás
    1.08
     keben
    1.05
     maler
    1.03
     dié
    1.02
     각
    1.00
     💕
    0.98
     angen
    0.98
     knap
    0.98
    Activations Density 0.028%

    No Known Activations