© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 31861
    Prev
    Next
    INDEX
    Explanations

    technical terms and non-english words

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

    This neuron responds to technical code tokens and literals—especially punctuation-heavy symbols (like regex delimiters “/.” or the degree sign “°”), uppercase identifiers (e.g. SYSTEM, OUT, Console), data-type names (integer, varchar), and other programming-related string literals.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    left
    -0.77
    Poppins
    -0.77
    olol
    -0.72
     الاجتما
    -0.71
    ropical
    -0.70
    zep
    -0.69
    uchter
    -0.69
    WISE
    -0.68
    Vigil
    -0.68
     ater
    -0.68
    POSITIVE LOGITS
     אפשר
    0.76
     kamar
    0.73
     žaid
    0.70
     sốt
    0.68
     specie
    0.68
     étudiant
    0.68
     Pong
    0.67
     味噌
    0.67
    ğın
    0.66
     neuroscience
    0.66
    Activations Density 0.025%

    No Known Activations