© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 33265
    Prev
    Next
    INDEX
    Explanations

    places and institutions

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

    This neuron fires on section‐style heading tokens—numbers, uppercase acronyms, and capitalized title words that typically appear at the start of report or article headings.

    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
    wijl
    -1.22
     工房
    -1.15
    AKI
    -1.14
     เรา
    -1.13
     Lorsqu
    -1.12
     lemn
    -1.09
    繙
    -1.09
     撮り
    -1.09
     negatif
    -1.09
     火鍋
    -1.08
    POSITIVE LOGITS
    :
    1.60
     about
    1.27
     not
    1.11
     using
    1.11
    ;
    1.09
     various
    1.02
     even
    1.02
     through
    1.00
    確認ください
    1.00
    并不
    0.97
    Activations Density 0.131%

    No Known Activations