© 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. 13079
    Prev
    Next
    INDEX
    Explanations

    additional funding or breakdown

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

    The neuron fires strongly on the very first content word of each new paragraph, effectively marking paragraph beginnings.

    oai_token-act-pair · o4-miniTriggered 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
    ”，“
    0.62
     Makarna
    0.60
     ル
    0.55
     ৷
    0.55
    ートル
    0.55
     バック
    0.55
     ​​​​
    0.54
    ْر
    0.53
     క్రికె
    0.52
    raseña
    0.52
    POSITIVE LOGITS
     مقاله
    1.55
     artikel
    1.47
     article
    1.42
     artikkel
    1.42
    篇文章
    1.38
    这篇文章
    1.38
    の記事
    1.34
    記事
    1.33
     статью
    1.30
     প্রবন্ধে
    1.28
    Activations Density 1.573%

    No Known Activations