© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 4714
    Prev
    Next
    INDEX
    Explanations

    websites and technical names

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

    The neuron activates on metadata and label tokens—especially proper‐name tags like place names (e.g. “Dubai,” “UAE,” “Saudi Arabia”), archive/category headings, and similar site‐generated labels.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    metadata and structural markers in blog posts and web content, particularly post timestamps, categories, and tags.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    webpage/blog metadata and navigation snippets indicating posting information like dates, authorship, categories/tags, archives, and RSS/feed references.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    trivial
    0.58
    两种
    0.57
    hepatic
    0.57
    メソッド
    0.55
    pyrim
    0.54
    impurity
    0.54
     uniformity
    0.53
    TLV
    0.53
    opter
    0.53
    ;'
    0.53
    POSITIVE LOGITS
     Blog
    0.97
     BLOG
    0.94
     Официа
    0.85
     blog
    0.83
     WordPress
    0.83
     Blogs
    0.82
     News
    0.80
     Consultado
    0.79
     Archive
    0.79
     blogging
    0.78
    Activations Density 0.995%

    No Known Activations