© 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. 58395
    Prev
    Next
    INDEX
    Explanations

    Roman numeral I

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

    The neuron fires on isolated single-letter tokens (especially the standalone “I” or “l”).

    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
     then
    -1.91
    if
    -1.84
     ensuring
    -1.61
     perpetually
    -1.59
     because
    -1.59
     unparalleled
    -1.59
    푣
    -1.58
     increases
    -1.55
    :
    -1.55
    An
    -1.52
    POSITIVE LOGITS
    ﹍﹍
    2.00
    鹟
    1.96
    嵛
    1.95
    oinette
    1.91
    힛
    1.90
    蓠
    1.87
    despite
    1.82
    鸲
    1.78
    
    1.76
     It
    1.75
    Activations Density 0.045%

    No Known Activations