© 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. 34-GEMMASCOPE-RES-131K
    7. 2839
    Prev
    Next
    INDEX
    Explanations

    invention and claims

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

    The neuron detects patent-style boilerplate reference phrases, e.g. “according to the present invention,” “of the preamble of claim,” and similar claim/invention citations.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    otheby
    -0.98
     kaksi
    -0.90
     ısı
    -0.88
    Ordin
    -0.85
     ním
    -0.84
    逻
    -0.83
     conhecido
    -0.82
     füh
    -0.82
    昔
    -0.82
     Streifen
    -0.81
    POSITIVE LOGITS
     according
    1.52
     claim
    1.50
     of
    1.09
     any
    1.05
     Claim
    0.95
     claims
    0.94
    According
    0.94
    によると
    0.90
    缺点
    0.90
     said
    0.86
    Activations Density 0.058%

    No Known Activations