© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 5850
    Prev
    Next
    INDEX
    Explanations

    Satya Nadella, Nadine, Nadia

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

    This neuron identifies blog‐post metadata phrases indicating where and when a post first appeared (e.g., “appeared first on …”).

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

    the phrase "appeared first on" in blog post footers that indicate content syndication or cross-posting.

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

    boilerplate blog attribution lines that state a post “appeared first on” a specific site or publication.

    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_31_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
     而
    0.78
    而
    0.76
    而在
    0.66
     mà
    0.60
    ಿಸ್
    0.58
    ERR
    0.58
     mentre
    0.58
     edu
    0.57
     decryption
    0.57
     mientras
    0.57
    POSITIVE LOGITS
    ))^
    0.54
    امت
    0.53
     アンティーク
    0.53
     dependent
    0.52
    طاء
    0.52
     Blogs
    0.51
     Dependent
    0.50
     dependant
    0.50
    ريس
    0.49
    protective
    0.49
    Activations Density 0.006%

    No Known Activations