© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 106424
    Prev
    Next
    INDEX
    Explanations

    by surprise

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

    The neuron triggers on the verb “take” (in its various inflected forms like taking, taken, took).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     säga
    -0.92
     sätta
    -0.91
     arbeta
    -0.88
     hjälpa
    -0.87
     försö
    -0.83
    UTHER
    -0.82
     använda
    -0.82
     tänka
    -0.81
     lägga
    -0.80
     börja
    -0.75
    POSITIVE LOGITS
     by
    3.36
     aback
    2.30
     surprise
    1.52
    surprise
    1.39
     By
    1.30
    By
    1.27
    by
    1.17
     BY
    1.08
     surpresa
    1.05
    โดย
    1.04
    Activations Density 0.008%

    No Known Activations