© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 100360
    Prev
    Next
    INDEX
    Explanations

    component inside folder

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

    This neuron activates on Chinese (and other CJK) character sequences, i.e. chunks of text in Chinese/Japanese script.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    多次
    -1.48
     blant
    -1.45
    ykjav
    -1.45
    もありました
    -1.38
    нями
    -1.35
    quería
    -1.34
     fleste
    -1.30
     like
    -1.29
     consecutively
    -1.27
     altres
    -1.26
    POSITIVE LOGITS
    マジで
    1.66
    ↵↵
    1.62
     YOUR
    1.56
    这人
    1.55
    獗
    1.48
    ほんと
    1.43
     дори
    1.41
    悲しい
    1.40
    かっこ
    1.38
    不愧
    1.38
    Activations Density 0.019%

    No Known Activations