© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 68521
    Prev
    Next
    INDEX
    Explanations

    code syntax punctuation

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

    Neuron 4 strongly activates on tokens that aren’t ordinary words but carry code-metadata syntax—IDs, version numbers, punctuation-heavy strings, and other symbol-laden identifiers.

    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
     sababu
    -0.90
    occuper
    -0.84
     teilweise
    -0.82
     dasselbe
    -0.82
     aéreo
    -0.81
    铆
    -0.81
     مستقیم
    -0.81
     atuação
    -0.81
     sessão
    -0.80
     Apesar
    -0.80
    POSITIVE LOGITS
    Fungsi
    1.02
    Özel
    0.95
    MenuGroup
    0.94
    Keine
    0.94
    Ayrıca
    0.92
    Portály
    0.92
    Açık
    0.91
     сент
    0.87
    gling
    0.86
    /></
    0.86
    Activations Density 0.077%

    No Known Activations