© 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. 128879
    Prev
    Next
    INDEX
    Explanations

    The neuron is specialized in picking up the numeric counts in “Category … :number” lines (the category identifier numbers at the end of a Wikipedia‐style entry).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    Hong Kong, Russian, Rugby, Actuaries

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     inget
    -1.09
     PROCEDURES
    -1.09
     grup
    -1.08
    İL
    -1.08
     Meksi
    -1.05
     olay
    -1.05
     Novembre
    -1.04
     histori
    -1.04
     haft
    -1.02
     katt
    -1.01
    POSITIVE LOGITS
     and
    1.30
    During
    1.02
     africaine
    1.01
    --
    0.93
     cercanas
    0.92
     awe
    0.91
    怎樣
    0.90
    主な
    0.89
    After
    0.88
     unseres
    0.85
    Activations Density 0.043%

    No Known Activations