© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 127786
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on tokens that are used as field or attribute names in structured data (e.g. XML/HTML/YAML keys before a colon or in tag attributes).

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408

    code elements and attributes

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    Negative Logits
     is
    -1.97
     will
    -1.70
     can
    -1.60
    T
    -1.52
    R
    -1.52
    P
    -1.52
    V
    -1.48
    l
    -1.48
     has
    -1.48
    to
    -1.48
    POSITIVE LOGITS
     nôtre
    1.84
     incrí
    1.77
     vôtre
    1.77
     État
    1.73
     fidé
    1.72
     évé
    1.71
     camiset
    1.69
     菇
    1.69
     blé
    1.64
    ܖ
    1.64
    Activations Density 0.012%

    No Known Activations