© 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. 72698
    Prev
    Next
    INDEX
    Explanations

    nineteen, untold story

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

    The neuron responds strongly to the word “nineteen” (as in years, centuries, or ages), i.e. spelled-out references to 19-xx dates or intervals.

    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
     is
    -3.44
    ↵↵
    -2.44
    驺
    -2.36
     compete
    -2.28
    に座
    -2.13
     introduce
    -2.09
     him
    -2.06
    かかる
    -2.06
     &
    -2.03
    ⢓
    -2.03
    POSITIVE LOGITS
    $
    3.17
     hervorge
    2.80
    嵛
    2.67
     OGSÅ
    2.63
    骝
    2.55
    堙
    2.39
    cakup
    2.30
     terlaris
    2.30
    骈
    2.25
    𖣘
    2.25
    Activations Density 0.002%

    No Known Activations