© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129597
    Prev
    Next
    INDEX
    Explanations

    The neuron is looking for punctuation and other non-word tokens (e.g. dashes, brackets, commas, acronyms) that typically appear in code or technical snippets.

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

    Show MPI radio technique

    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
    我们
    -1.06
    老
    -0.98
     jeżeli
    -0.94
    我们的
    -0.94
    不能
    -0.94
     amely
    -0.92
    sauce
    -0.92
    救
    -0.91
    ).—
    -0.90
    很
    -0.89
    POSITIVE LOGITS
     χρησιμοποι
    0.96
    actéristique
    0.96
     nació
    0.95
     étudier
    0.95
    Kary
    0.93
    wyczaj
    0.90
    在场
    0.89
    Waarom
    0.88
     geringen
    0.88
    titura
    0.87
    Activations Density 0.147%

    No Known Activations