© 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. 10-GEMMASCOPE-RES-131K
    7. 72270
    Prev
    Next
    INDEX
    Explanations

    GPX, RGDX, mEPHX, JSX

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

    The neuron detects the letter "X" in technological contexts, particularly in programming and scientific notation.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron reliably activates on occurrences of the token “X” when it’s part of “JSX” (i.e. it flags mentions or code of JSX).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     comprising
    -2.94
    柃
    -2.52
    .";
    -2.47
     that
    -2.44
     includes
    -2.44
     achieving
    -2.39
    یب
    -2.39
    。”
    -2.39
     which
    -2.36
    n
    -2.30
    POSITIVE LOGITS
     คน
    2.83
    
    2.59
    厹
    2.55
    1
    2.53
     REALLY
    2.44
     trabajando
    2.39
    で
    2.34
     шесть
    2.33
     ส์
    2.30
    Simply
    2.28
    Activations Density 0.008%

    No Known Activations