© 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. 22470
    Prev
    Next
    INDEX
    Explanations

    numbers preceded by "no." or "number"

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

    The neuron identifies standalone numeric strings—especially volume, issue, and page numbers in bibliographic citations.

    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
     who
    -2.00
     all
    -1.72
     what
    -1.65
    ：……
    -1.64
     only
    -1.58
     toată
    -1.54
    attributed
    -1.52
     use
    -1.51
     from
    -1.49
     same
    -1.48
    POSITIVE LOGITS
     rojas
    1.85
     membersihkan
    1.77
     „
    1.72
     memperbaiki
    1.66
    llten
    1.59
     („
    1.57
    naires
    1.55
    柃
    1.52
     hadde
    1.48
     merawat
    1.48
    Activations Density 0.026%

    No Known Activations