© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 76389
    Prev
    Next
    INDEX
    Explanations

    patent numbers and dates

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

    The neuron fires on numeric tokens (digits and slashes) typically found in reference numbers, dates, and patent or model designations.

    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
     doesn
    -0.97
     didn
    -0.90
     don
    -0.84
     gammel
    -0.76
    Justice
    -0.74
    abras
    -0.73
     wouldn
    -0.73
     Wouldn
    -0.72
     wasn
    -0.71
    hlungen
    -0.71
    POSITIVE LOGITS
     Banner
    0.83
    Vía
    0.81
    たび
    0.80
    Connecting
    0.75
     vez
    0.73
     traer
    0.72
     benötigt
    0.71
    Več
    0.71
    konen
    0.70
    〝
    0.69
    Activations Density 0.018%

    No Known Activations