© 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. 84451
    Prev
    Next
    INDEX
    Explanations

    multilingual suffixes

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

    The neuron fires on words that are built from a stem plus a common suffix (e.g. “tint-ing,” “hand-outs,” “fore-casts,” “John-son”), i.e. it detects stem+suffix word formations.

    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
    んでる
    -1.25
    lasyon
    -1.13
    Gilla
    -1.08
    UTION
    -1.05
    ltä
    -1.05
     neighborhood
    -1.04
    ẕ
    -1.04
    ksiyon
    -1.04
     favorable
    -1.03
     stasiun
    -1.02
    POSITIVE LOGITS
    Seguro
    1.38
     lixo
    1.33
     souple
    1.32
    Flere
    1.26
     álcool
    1.22
     sombras
    1.20
    جوی
    1.18
     jaune
    1.18
     ны
    1.17
     cepillo
    1.17
    Activations Density 0.000%

    No Known Activations