© 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. 67787
    Prev
    Next
    INDEX
    Explanations

    roman numerals followed by context

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

    The neuron detects Roman‐numeral–style numbering (e.g. i, ii, vi, xi) and related formatting markers in document headers or tables of contents.

    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
     our
    -1.63
    a
    -1.59
     your
    -1.59
     about
    -1.53
     you
    -1.48
     it
    -1.48
     waarvan
    -1.47
     using
    -1.41
     employing
    -1.34
     alleging
    -1.30
    POSITIVE LOGITS
     где
    1.80
    4
    1.59
     solchen
    1.56
    dzić
    1.50
    replyreport
    1.47
    </h5>
    1.43
    孥
    1.42
    attva
    1.41
    閨
    1.41
     ۱۸
    1.34
    Activations Density 0.012%

    No Known Activations