© 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. 10-GEMMASCOPE-RES-131K
    7. 86560
    Prev
    Next
    INDEX
    Explanations

    legislator

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

    The neuron fires on rare or specialized subwords—proper names, foreign terms, accented words, or other low‐frequency tokens.

    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
    塍
    -2.58
    因为它
    -2.45
    ↵
    -2.42
    比如说
    -2.31
     fãs
    -2.22
    认可
    -2.17
    我们会
    -2.17
    俱乐
    -2.16
    請繼續
    -2.13
    皞
    -2.09
    POSITIVE LOGITS
    ’
    3.72
    </i>
    2.92
    ン
    2.91
    н
    2.86
    과
    2.84
    The
    2.67
    could
    2.53
    った
    2.52
    1
    2.52
    How
    2.50
    Activations Density 0.007%

    No Known Activations