© 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. 10-GEMMASCOPE-RES-131K
    7. 104484
    Prev
    Next
    INDEX
    Explanations

    na followed by various words

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

    The neuron consistently fires on the two-letter string “na” (e.g. as a standalone token or part of words), regardless of language or context.

    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
    {
    -2.70
    e
    -2.67
    紧接着
    -2.36
    ians
    -2.25
    ſſen
    -2.19
     then
    -2.19
     CommandType
    -2.05
    ó
    -2.00
    ’
    -2.00
    o
    -2.00
    POSITIVE LOGITS
     he
    2.31
    However
    2.30
     a
    2.16
    
    2.06
    тся
    1.98
    ふう
    1.98
    狨
    1.95
     werden
    1.91
     Söz
    1.91
    They
    1.88
    Activations Density 0.001%

    No Known Activations