© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 5472
    Prev
    Next
    INDEX
    Explanations

    This neuron fires on the first significant (usually capitalized or content-heavy) word at the start of a new paragraph or block, effectively marking headings or section-openers.

    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
    歓迎
    0.79
     domande
    0.78
    オ
    0.77
     Lombardi
    0.75
    ară
    0.75
    オフ
    0.74
     közel
    0.73
    から
    0.73
    質問
    0.71
    ブ
    0.70
    POSITIVE LOGITS
     применя
    0.91
     coverings
    0.81
     выпуска
    0.79
     swings
    0.78
    riage
    0.77
     flicks
    0.77
     данного
    0.75
    riages
    0.74
    чены
    0.74
     каждого
    0.74
    Activations Density 0.000%

    No Known Activations