© 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. 12876
    Prev
    Next
    INDEX
    Explanations

    notes at the end

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

    This neuron strongly activates on common nouns that serve as the objects of prepositional phrases (e.g., country, ships, room, online, morning, phone, world, life).

    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
    ۔
    0.68
    ។
    0.67
     silam
    0.62
    .。
    0.62
    érica
    0.59
    埼玉
    0.59
    ombasa
    0.58
    .`
    0.57
    umberland
    0.57
    landa
    0.57
    POSITIVE LOGITS
     nejsou
    0.64
     kunnen
    0.60
     используются
    0.60
     существуют
    0.59
     animals
    0.57
    系统中
    0.56
     various
    0.55
     могат
    0.54
     aren
    0.54
    ните
    0.54
    Activations Density 0.561%

    No Known Activations