© 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. 22-GEMMASCOPE-RES-131K
    7. 128818
    Prev
    Next
    INDEX
    Explanations

    The neuron detects technical or specification field labels (e.g. “Type:”, “Version:”, “Dimension:”, etc.) in structured document metadata.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    Type status description

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     a
    -1.48
     an
    -1.43
     that
    -1.38
     any
    -1.29
     their
    -1.28
     daarom
    -1.27
     With
    -1.25
     !!
    -1.24
     our
    -1.17
     !!!
    -1.16
    POSITIVE LOGITS
    は
    1.45
     äldre
    1.39
    5
    1.37
    7
    1.37
    <bos>
    1.36
     handels
    1.32
    6
    1.32
    かというと
    1.31
    Manfaat
    1.30
    があるので
    1.29
    Activations Density 0.002%

    No Known Activations