© 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. 10-GEMMASCOPE-RES-131K
    7. 15179
    Prev
    Next
    INDEX
    Explanations

    introducing an exception or detail

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

    This neuron detects formal clause-introducing terms (especially “wherein”) common in patent-style technical descriptions.

    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
    );
    -2.14
    ,”
    -2.05
    ”、“
    -1.92
    ungeon
    -1.90
     contado
    -1.88
     tevé
    -1.78
    "]="
    -1.70
     cortada
    -1.65
     a
    -1.65
    embroidered
    -1.64
    POSITIVE LOGITS
     they
    1.95
     there
    1.88
     dieß
    1.73
     demikian
    1.70
     that
    1.66
     Bibl
    1.65
    𝅥
    1.63
     it
    1.62
     Akku
    1.59
     EACH
    1.55
    Activations Density 0.015%

    No Known Activations