© 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. 59203
    Prev
    Next
    INDEX
    Explanations

    XML namespaces and document formats

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

    This neuron is strongly activated by runs of whitespace characters (e.g. leading spaces or tabs), effectively detecting indented or blank regions in the text.

    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
     rodríguez
    -1.76
    珦
    -1.73
    pullover
    -1.65
     gonzález
    -1.59
     what
    -1.53
    睒
    -1.53
     industrias
    -1.52
     garcía
    -1.52
     sánchez
    -1.51
    して下さい
    -1.49
    POSITIVE LOGITS
     others
    1.77
    .
    1.47
    like
    1.46
    を読む
    1.43
     adresses
    1.33
    such
    1.30
     autorisation
    1.30
     Remains
    1.30
    ous
    1.27
     oth
    1.26
    Activations Density 0.006%

    No Known Activations