© 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. 100496
    Prev
    Next
    INDEX
    Explanations

    sending links or content

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

    Chinese characters and text, particularly in technical or programming contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron primarily fires on Chinese words (i.e. it detects Chinese‐language tokens).

    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
    糨
    -2.81
    媜
    -2.75
    馀
    -2.70
     haja
    -2.69
    だろうか
    -2.69
    揠
    -2.66
    which
    -2.64
    迓
    -2.63
    艄
    -2.61
    ICOLON
    -2.50
    POSITIVE LOGITS
    此时
    3.63
    1
    3.08
     Lokasi
    2.45
     Sebuah
    2.39
     Perubahan
    2.38
     ailing
    2.36
     Masalah
    2.36
     Welche
    2.31
     Often
    2.30
     Replacing
    2.30
    Activations Density 0.002%

    No Known Activations