© 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. 58409
    Prev
    Next
    INDEX
    Explanations

    multi-part or part + descriptor

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

    The neuron strongly activates on occurrences of the word “part” (as in “Part I,” “two-part,” “multi-part,” etc.), marking text that labels sections of a multi-part series.

    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
    {
    -2.30
    '
    -1.80
     &
    -1.60
    们的
    -1.54
     brief
    -1.49
     carcasa
    -1.46
    .
    -1.45
     番目
    -1.44
    蛲
    -1.43
    -1.42
    POSITIVE LOGITS
    on
    2.16
    hline
    1.98
    u
    1.91
    棪
    1.87
    ar
    1.86
    ”
    1.84
    }{
    1.70
    栻
    1.69
    section
    1.69
     We
    1.66
    Activations Density 0.008%
    No Known Activations