© 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. 22-GEMMASCOPE-RES-131K
    4. 13036
    Prev
    Next
    INDEX
    Explanations

    In practice, Entries from, Ground Zero

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

    The neuron fires on tokens that are being used as labels or identifiers—e.g. XML `<key>` tag names in plist files and named code entities like function or property identifiers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     only
    -1.20
    larını
    -0.96
     "",
    -0.96
     Json
    -0.88
     that
    -0.86
    },
    
    -0.85
     most
    -0.84
    学家
    -0.84
    ",
    -0.84
     ",
    -0.84
    POSITIVE LOGITS
    entas
    0.95
    diyesi
    0.94
    如图
    0.94
     Danes
    0.94
    印刷
    0.93
     resim
    0.93
    cuál
    0.88
    gnant
    0.88
     asl
    0.87
     Persia
    0.87
    Activations Density 0.001%

    No Known Activations