© 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. 73651
    Prev
    Next
    INDEX
    Explanations

    code delimiters

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

    The neuron fires on literal delimiter characters—especially quotation marks, braces, and bracket symbols—commonly used in code (e.g. regex boundaries or quoted strings).

    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
     این
    -1.28
     This
    -1.27
     diesem
    -1.20
     Weltkrieg
    -1.16
     этом
    -1.15
     этой
    -1.12
     där
    -1.10
     гиб
    -1.09
    thansa
    -1.09
     هذا
    -1.07
    POSITIVE LOGITS
    かなかった
    1.18
    但是在
    1.15
    infty
    1.12
     confines
    1.11
     comment
    1.07
     χρησιμοποι
    1.06
     customer
    1.02
     they
    1.02
    中には
    1.01
     сайты
    0.98
    Activations Density 0.018%

    No Known Activations