© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 102854
    Prev
    Next
    INDEX
    Explanations

    file paths, dates, and numbers

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

    The neuron is detecting “structural” tokens and delimiters often used in code or markup (e.g. angle‐brackets, commas, stars, keywords like header/None/PUBLIC/SPECIAL).

    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
    纯
    -0.69
    Keywords
    -0.68
    貴重
    -0.67
    顛
    -0.66
    Target
    -0.65
     дов
    -0.64
     inadequate
    -0.62
     triggered
    -0.62
    ᅪ
    -0.61
     desped
    -0.61
    POSITIVE LOGITS
    HCI
    0.87
    ੍ਰ
    0.75
    ofer
    0.75
     RY
    0.73
     Roh
    0.72
     газе
    0.72
     PCI
    0.71
     inan
    0.71
     seo
    0.70
    NSE
    0.70
    Activations Density 0.073%

    No Known Activations