© 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. 128278
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on text inside parentheses or similar bracketing, effectively detecting parenthetical segments.

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

    recent advances

    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
     their
    -1.23
     revisor
    -1.21
     diretor
    -1.14
     miliar
    -1.13
     inova
    -1.11
    idade
    -1.10
     Temporada
    -1.09
     profesor
    -1.09
     autentic
    -1.09
    findViewById
    -1.07
    POSITIVE LOGITS
     generazione
    1.13
    licans
    0.97
     wou
    0.97
    exprimer
    0.96
    roleum
    0.95
     ashes
    0.94
     comunicarse
    0.93
    なあ
    0.93
     provenientes
    0.91
    utenant
    0.90
    Activations Density 0.003%

    No Known Activations