© 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. 128566
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on long runs of whitespace or indentation (i.e. lines composed mostly of spaces).

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

    no or want

    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
     if
    -1.47
     from
    -1.42
    sprzed
    -1.39
     berangkat
    -1.30
    archiw
    -1.28
    Czytaj
    -1.24
     dictionnaire
    -1.24
    angkap
    -1.23
    ꉂ
    -1.20
     menik
    -1.19
    POSITIVE LOGITS
    ak
    1.45
    ar
    1.42
    |
    1.41
    o
    1.40
    na
    1.38
    ne
    1.37
    un
    1.35
    Y
    1.34
    an
    1.30
    el
    1.30
    Activations Density 0.000%

    No Known Activations