© 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. 22-GEMMASCOPE-RES-131K
    7. 16432
    Prev
    Next
    INDEX
    Explanations

    modifying contextual terms

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

    The neuron fires most strongly on mid-frequency technical or scholarly content words (e.g. “philosophical,” “research,” “methods,” “characterize”), effectively detecting academic/scientific jargon rather than common function words or punctuation.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     September
    -0.80
     Because
    -0.77
    ragamo
    -0.76
     пун
    -0.76
    אָ
    -0.76
    出了
    -0.76
     Have
    -0.75
    垂
    -0.75
     osserva
    -0.75
     terrible
    -0.73
    POSITIVE LOGITS
     frein
    0.97
    $\
    0.88
    entraînement
    0.86
     zoals
    0.80
     werk
    0.76
     witte
    0.76
    paw
    0.75
     $_{
    0.75
     penser
    0.75
     orale
    0.75
    Activations Density 3.269%

    No Known Activations