© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 15090
    Prev
    Next
    INDEX
    Explanations

    parts and systems

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

    This neuron detects the formal section‐heading and boilerplate phrases that appear in patent documents (e.g. “Field of the Invention,” “The present invention relates to,” numbered sections, etc.).

    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
    xii
    -1.70
    ٨
    -1.63
    xiii
    -1.62
    viii
    -1.60
    ٦
    -1.57
    ٣
    -1.55
     this
    -1.54
    ,”
    -1.52
    Otra
    -1.51
    ̶
    -1.51
    POSITIVE LOGITS
    However
    1.54
     but
    1.51
    acier
    1.45
    достатки
    1.41
     \'
    1.39
    Jawab
    1.35
    Salu
    1.35
     citazioni
    1.30
    ・・・・・・・・
    1.30
    옙
    1.30
    Activations Density 32.636%

    No Known Activations