© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 32130
    Prev
    Next
    INDEX
    Explanations

    word + up/over/off

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

    This neuron fires on the first element of hyphenated compounds (e.g. “cover-up,” “roll-out,” “follow-up,” “build-up,” “change-over”).

    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
     to
    -0.84
     man
    -0.81
    erobic
    -0.80
    羔
    -0.80
    лади
    -0.79
    cended
    -0.79
    itanya
    -0.78
    ‌هاي
    -0.77
    слови
    -0.76
    asional
    -0.76
    POSITIVE LOGITS
    up
    1.36
    ups
    1.27
     ups
    1.09
    -
    0.93
    為に
    0.85
    off
    0.84
    offs
    0.84
    tup
    0.83
    out
    0.81
    Ups
    0.80
    Activations Density 0.053%

    No Known Activations