© 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. 113804
    Prev
    Next
    INDEX
    Explanations

    URLs, copyright, discounts, questions

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

    whitespace and formatting characters, particularly tabs and multiple spaces used for indentation or alignment in structured documents.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron detects tab characters (indentation tabs) 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
    s
    -2.64
     hygienic
    -1.95
    es
    -1.93
    艰难
    -1.90
     prevents
    -1.84
     oftentimes
    -1.82
    ógico
    -1.80
    ⸸
    -1.77
     invigor
    -1.77
    üşonlu
    -1.77
    POSITIVE LOGITS
    \
    2.13
    䛗
    2.11
    the
    2.06
    PLEASE
    2.00
    .”
    1.95
    鲩
    1.95
    while
    1.79
    ’
    1.73
    like
    1.73
     kapture
    1.72
    Activations Density 0.007%

    No Known Activations