© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 38423
    Prev
    Next
    INDEX
    Explanations

    Hawaii, AG, Order, Election, Chinatown

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

    The neuron is tuned to pick out structural or metadata tokens—like headings, section numbers, official names, and other capitalized identifiers or numbers indicating document organization.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     dolay
    -1.09
    ganz
    -1.01
     dinosaurio
    -1.00
    ָד
    -0.99
    História
    -0.97
    Consejos
    -0.96
    Artículo
    -0.95
    enseur
    -0.95
     kulak
    -0.94
     倉
    -0.94
    POSITIVE LOGITS
    łoń
    1.14
     isn
    1.09
     Tân
    1.08
     schonmal
    1.07
    COVID
    1.03
    [
    1.03
     ain
    1.02
     Ian
    1.00
     COVID
    0.99
    などなど
    0.98
    Activations Density 0.102%

    No Known Activations