© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26
    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    2. Qwen3-32B
    3. MLP Neurons - Layer 3
    4. 21737
    Prev
    Next
    Layer 2Layer 4
    INDEX
    Explanations

    Chinese 年 (year) compounds

    np_max-act · claude-sonnet-5-5
    Triggered by @paultiplady

    characters following 年

    np_max-act-logits · claude-sonnet-5-5
    Triggered by @paultiplady

    少

    np_min-act-logits · claude-sonnet-5-5Neg Acts
    Triggered by @paultiplady

    Chinese/Japanese tokens that complete a fixed term tied to the year-end or New Year period or to annual events, such as the final character of words for New Year goods, year-end meals, or yearly chronicles.

    oai_token-act-pair · claude-sonnet-5-5
    Triggered by @paultiplady

    The token that completes a Chinese or Japanese compound word containing 年 ("year"), typically the character right after 年 or a broken byte fragment of it. These compounds are mostly about new year, calendars and annual events, such as 年貨, 年菜, 年節, 年玉, 年式, 萬年曆, 編年史 and 年夜飯.

    eleuther_acts_top20 · claude-sonnet-5-5
    Triggered by @paultiplady
    New Auto-Interp
    Top Features by Cosine Similarity
    Embeds
    IFrame
    Link
    Loading dashboard...