© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 60375
    Prev
    Next
    INDEX
    Explanations

    Zea mays

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

    sequences of "ea" followed by numbers in code or hexadecimal strings.

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

    The neuron strongly activates on the letter sequence “ea” wherever it appears.

    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
    進入
    -2.48
    龊
    -2.47
    But
    -2.45
    ”,
    -2.44
    ).
    -2.44
     “
    -2.34
    It
    -2.34
    The
    -2.34
    
    -2.33
    In
    -2.31
    POSITIVE LOGITS
    ea
    3.84
    튿
    3.14
     are
    3.13
    geſ
    2.64
    
    2.50
    ⵈ
    2.48
    ?"
    2.45
     we
    2.44
    {
    2.44
     Monfieur
    2.39
    Activations Density 0.003%

    No Known Activations