© 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. 34-GEMMASCOPE-RES-131K
    7. 2127
    Prev
    Next
    INDEX
    Explanations

    specific contexts or categorizations

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

    This neuron selectively activates on Japanese-language tokens (e.g. Kanji or common Japanese words), marking segments of text written in Japanese.

    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
     yesterday
    -1.02
    viembre
    -1.02
    豊富な
    -1.00
     प्रत्येक
    -0.96
    ])
    -0.93
    什么呢
    -0.91
    asztal
    -0.91
    vky
    -0.90
    asında
    -0.90
     divisions
    -0.90
    POSITIVE LOGITS
    의
    1.38
    부터
    1.20
    、
    1.02
     olduğ
    0.97
    ,
    0.93
     نیز
    0.91
     particularly
    0.90
     like
    0.90
    ，
    0.88
    indest
    0.87
    Activations Density 0.001%

    No Known Activations