© 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. 39619
    Prev
    Next
    INDEX
    Explanations

    Elect following prefix + word segmentation

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

    The neuron detects fragments of web URLs, especially protocol separators ("://"), domain dots ("."), top-level domains (like "com"), and other URL-specific punctuation.

    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
    rativa
    -1.15
     nedan
    -1.14
     determina
    -1.14
    ってください
    -1.14
    と呼ば
    -1.10
     they
    -1.05
    MeasureSpec
    -1.05
    ($__
    -1.03
     formule
    -1.03
    zett
    -1.02
    POSITIVE LOGITS
     poderá
    1.23
    However
    1.19
     mantend
    1.18
     podrá
    1.17
     gostar
    1.16
     aproveitar
    1.16
    Puedes
    1.14
    くださいね
    1.13
     presentará
    1.11
    욤
    1.11
    Activations Density 0.000%

    No Known Activations