© 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. 128894
    Prev
    Next
    INDEX
    Explanations

    This neuron detects tokens that come from code or markup snippets (e.g. programming identifiers, import paths, HTML tags).

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

    programming dependency

    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
     tica
    -1.26
     ananas
    -1.04
    İL
    -1.01
    üks
    -1.01
    dafri
    -0.97
     camion
    -0.97
     spoor
    -0.97
     spesies
    -0.96
     cruce
    -0.96
    hadoop
    -0.96
    POSITIVE LOGITS
     ...
    1.21
    <eos>
    1.15
     ...)
    1.08
     [...]
    0.95
    การณ์
    0.92
    [...]
    0.90
    ...
    0.89
     contex
    0.86
     powy
    0.85
     ...]
    0.84
    Activations Density 0.022%

    No Known Activations