© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 13326
    Prev
    Next
    INDEX
    Explanations

    библиотека, principality, standard

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

    The neuron is sensitive to tokens written in non-English scripts or with non-ASCII characters (foreign-language or accented words).

    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
    Only
    -1.41
    って思
    -1.36
     Only
    -1.36
    signable
    -1.31
    もありますが
    -1.27
    mayın
    -1.18
    veled
    -1.18
     täv
    -1.16
    四个
    -1.15
     at
    -1.13
    POSITIVE LOGITS
     of
    2.42
     ángel
    1.37
     cinturón
    1.33
     chaleco
    1.30
    UAWEI
    1.27
     anillo
    1.27
    ときは
    1.24
     obrázok
    1.22
     gorro
    1.20
     Belgische
    1.19
    Activations Density 0.030%

    No Known Activations