© 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. 38031
    Prev
    Next
    INDEX
    Explanations

    file headers or descriptions

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

    The neuron activates strongly on technical metadata labels and numeric or version‐style tokens (e.g. “description,” “Name,” “Comment,” “Exec,” numeric version strings, and constants).

    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
    hôpital
    -0.70
     number
    -0.68
     handling
    -0.65
    grunt
    -0.65
     Number
    -0.65
     type
    -0.65
    IVAS
    -0.65
    ੁ
    -0.64
    ATV
    -0.64
    NUMBER
    -0.64
    POSITIVE LOGITS
    km
    0.67
    庆
    0.66
    接受
    0.65
    Megan
    0.64
     Pd
    0.64
    Spre
    0.63
    hua
    0.63
    DateFormat
    0.62
    frequency
    0.61
     三
    0.61
    Activations Density 0.094%

    No Known Activations