© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 73939
    Prev
    Next
    INDEX
    Explanations

    cancer predisposition syndromes

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

    horizontal dividing lines made of dashes and other special characters.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron flags unusually long single tokens—whether long words (e.g. “ArgumentNullException,” “predisposition”) or long runs of punctuation (e.g. “-------------”).

    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
    t
    -2.69
    When
    -2.48
     When
    -2.45
    Another
    -2.34
    the
    -2.28
    These
    -2.27
    Every
    -2.25
    Despite
    -2.22
     装修
    -2.20
    From
    -2.16
    POSITIVE LOGITS
    “
    3.05
     '
    2.73
     of
    2.69
     “
    2.41
    和她
    2.34
    ,“
    2.28
     they
    2.14
     mantra
    2.13
     a
    2.11
     largely
    2.06
    Activations Density 0.003%

    No Known Activations