© 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. 33417
    Prev
    Next
    INDEX
    Explanations

    xml code structures

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

    The neuron detects hexadecimal color‐code attributes (e.g. the “#FFFFFF” style color or bgcolor assignments) in markup tags.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     becomes
    -1.24
     how
    -1.23
     have
    -1.19
     as
    -1.10
     creating
    -1.09
     to
    -1.09
    はなく
    -1.08
     now
    -1.07
     produces
    -1.06
     from
    -1.04
    POSITIVE LOGITS
    〮
    1.22
    きれ
    1.22
    雱
    1.20
    ource
    1.20
    えっ
    1.16
    tablissements
    1.14
    rdoba
    1.14
     arī
    1.14
    Fuente
    1.13
    うわ
    1.13
    Activations Density 0.001%

    No Known Activations