© 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. 10-GEMMASCOPE-RES-131K
    7. 70771
    Prev
    Next
    INDEX
    Explanations

    foreign language verbs and nouns

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

    The neuron is keyed to stand-alone, title-style capitalized words (e.g. names, headings, place names)—tokens in Title Case or ALL CAPS that typically appear in document metadata or section titles.

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
     Abstr
    -2.06
     berge
    -1.88
     Janvier
    -1.77
     вернуться
    -1.77
     dij
    -1.72
     vanden
    -1.72
    ',
    -1.70
     Buku
    -1.70
     Novembre
    -1.67
     Maak
    -1.66
    POSITIVE LOGITS
     would
    2.05
     když
    1.95
     colheres
    1.86
     tornou
    1.86
     is
    1.84
     когато
    1.82
     obtend
    1.79
     permanecer
    1.77
     encontrará
    1.74
     мужская
    1.74
    Activations Density 0.177%

    No Known Activations