© 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. 127931
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on numeric literals (e.g. dimension or position values) embedded in code or markup.

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

    spatial positions

    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
     etc
    -1.18
     to
    -1.16
     versus
    -1.13
     into
    -1.10
     iyong
    -1.09
     such
    -1.09
     who
    -1.06
     will
    -1.05
      
    -1.02
     studied
    -0.98
    POSITIVE LOGITS
    舠
    1.13
    Lø
    1.13
     незавер
    1.11
    gól
    1.08
    ultu
    1.08
    ミアム
    1.07
     jaunes
    1.07
     tarjetas
    1.05
     пун
    1.02
    เกง
    1.02
    Activations Density 0.004%

    No Known Activations