© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 8099
    Prev
    Next
    INDEX
    Explanations

    Do followed by 'ane', 'ppler', 'zer', 'zier', 'M', 'It', 'Parallel', 'cervical'

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

    The neuron is highly sensitive to the exact token “Do” (an uppercase D followed by lowercase o).

    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
    viembre
    -2.86
    ⵌ
    -2.55
     stuks
    -2.52
     problemet
    -2.44
     mengungkapkan
    -2.33
     bús
    -2.30
    économ
    -2.27
     szczeg
    -2.23
     عندما
    -2.23
    ми
    -2.23
    POSITIVE LOGITS
     It
    2.91
     However
    2.80
     was
    2.73
     Whether
    2.72
    {
    2.64
    -
    2.36
    *
    2.28
     Its
    2.23
     About
    2.19
    随即
    2.17
    Activations Density 0.037%

    No Known Activations