© 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. 58410
    Prev
    Next
    INDEX
    Explanations

    use this

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

    The neuron fires on the token “use,” detecting occurrences of that word (as in “you may not use this file…” in license headers).

    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
    ’
    -3.70
    l
    -2.97
    ",
    -2.75
     they
    -2.64
    \
    -2.63
    櫺
    -2.50
    鍆
    -2.50
    縠
    -2.50
    n
    -2.44
    淠
    -2.38
    POSITIVE LOGITS
    เพราะ
    2.22
     embody
    2.19
    塤
    2.17
    了一个
    2.05
    懨
    2.03
    了一個
    1.98
    みたら
    1.95
    型の
    1.95
     見る
    1.94
     peč
    1.94
    Activations Density 0.298%

    No Known Activations