© 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. 22-GEMMASCOPE-RES-131K
    7. 33589
    Prev
    Next
    INDEX
    Explanations

    dates or numerical sequences

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

    The neuron strongly activates on very common short function words (e.g. “in,” “the,” “de,” “to”), i.e. stopwords.

    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
    総評
    -1.16
     seiz
    -1.14
     også
    -1.07
     メニュー
    -1.02
    コスプレ
    -1.01
    Ширина
    -1.01
    ハンドメイド
    -0.99
    踔
    -0.97
    小孩子
    -0.97
     rapids
    -0.96
    POSITIVE LOGITS
     other
    1.62
     verden
    1.05
    ctory
    1.00
    thodox
    0.99
    cology
    0.95
    Ļ
    0.94
     niego
    0.93
    berapa
    0.92
    romeda
    0.91
    監
    0.91
    Activations Density 0.001%

    No Known Activations