© 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. 64696
    Prev
    Next
    INDEX
    Explanations

    legal document separators

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

    This neuron detects unusually long runs of whitespace (large indentation or alignment blocks) in code or license headers.

    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
     (
    -1.59
     so
    -1.48
     in
    -1.45
     one
    -1.42
     After
    -1.30
     after
    -1.29
     all
    -1.27
     an
    -1.26
     most
    -1.26
     JMenuItem
    -1.25
    POSITIVE LOGITS
     volledig
    1.61
     vijf
    1.52
     prachtige
    1.48
     belangrijke
    1.42
     geweldige
    1.38
     feltro
    1.37
     voldoende
    1.36
     certamente
    1.36
     bijzondere
    1.30
     volledige
    1.30
    Activations Density 0.035%

    No Known Activations