© 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 59
    4. 11079
    Prev
    Next
    Layer 58Layer 60
    INDEX
    Explanations

    holiday Method used: 2. Most tokens after the max activating token are "holiday" or "holidays", with a few related words like "Christmas"

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

    Christmas/holiday upcoming

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

    Christmas/New Year holidays

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

    Christmas (Natale/Noël/Рождество) and Christmas-related holiday context in non-English text, especially the word for Christmas itself and the words just before it, such as "Christmas Eve", "Christmas day", or "New Year".

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

    Multilingual text (Italian, French, Russian, English) about Christmas and New Year holidays. The marked tokens are mostly function words, prepositions, commas, and word fragments that come right before or after a holiday name (Natale, Noël, Christmas, Capodanno, Рождество, Новый год). Examples are "di" in "vigilia di Natale", "on" in "on Christmas Day", "avant" in "avant Noël", and commas in lists of holidays. The pattern is the connecting grammar that links a time reference to a Christmas or New Year festivity.

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