© 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. 22-GEMMASCOPE-RES-131K
    7. 9454
    Prev
    Next
    INDEX
    Explanations

    testing and diagnostics

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

    error handling patterns in programming code, particularly when checking for errors and throwing exceptions.

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

    The neuron fires on error‐handling code patterns—specifically the “if (err) … throw err” style checks and throws.

    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
     nemlig
    -0.86
     Translated
    -0.86
     artikke
    -0.83
    lague
    -0.80
     compositeur
    -0.80
     Cholesterol
    -0.79
     derfor
    -0.78
     khuyên
    -0.77
    Virus
    -0.77
     يوسف
    -0.76
    POSITIVE LOGITS
     test
    1.24
    ۶
    1.13
     testing
    1.09
     emisiones
    1.09
    utschland
    1.08
     tecnologías
    1.00
    <>();
    1.00
     examine
    0.98
     testGet
    0.97
     pruebas
    0.96
    Activations Density 0.007%

    No Known Activations