© 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. 7175
    Prev
    Next
    INDEX
    Explanations

    pretend, understanding, her, popular

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

    The neuron fires on text that’s laying out procedures or “how-to” explanations—i.e. instructional or guideline-style passages.

    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_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
    با
    0.78
    не
    0.75
    ंसाठी
    0.75
    મ
    0.73
    由
    0.72
    Designed
    0.71
    нг
    0.71
    साठी
    0.70
    Illustrated
    0.70
    ς
    0.69
    POSITIVE LOGITS
     Urology
    0.89
     alkaloids
    0.88
     caud
    0.88
     duodenum
    0.87
     Gertrude
    0.86
     junctions
    0.85
     dấu
    0.84
     หน้า
    0.83
     khand
    0.82
     glutamate
    0.81
    Activations Density 0.000%

    No Known Activations