© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 129310
    Prev
    Next
    INDEX
    Explanations

    repulsion, part, possible, summary, department

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

    The neuron detects tokens that label structural divisions or sections of a document—e.g. words meaning “part,” “section,” or “overview.”

    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
     
    -3.08
    事も
    -2.56
    をし
    -2.52
     وأ
    -2.45
     Anſ
    -2.45
    太过
    -2.42
     لأنه
    -2.42
     وإ
    -2.42
    くらいの
    -2.39
     développ
    -2.36
    POSITIVE LOGITS
    n
    3.23
    晅
    2.89
    这是什么
    2.83
    '
    2.83
     was
    2.81
    en
    2.80
     شیپور
    2.67
    ly
    2.61
    re
    2.45
     estekak
    2.45
    Activations Density 0.001%

    No Known Activations