© 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. 31-GEMMASCOPE-2-RES-16K
    7. 10439
    Prev
    Next
    INDEX
    Explanations

    tailpiece, tangible

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

    The neuron spotlights subword tokens from proper names—especially capitalized names of people, places, events, or organizations.

    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
    cumin
    0.54
     STL
    0.52
    ੈ
    0.51
    人为
    0.50
     decent
    0.49
    ائمة
    0.49
    න්ත
    0.48
    achite
    0.47
     atteindre
    0.47
    TableHeader
    0.47
    POSITIVE LOGITS
    ekwondo
    0.83
    iseach
    0.64
    ächlich
    0.62
     タ
    0.62
    anyika
    0.61
     čin
    0.59
    क्सी
    0.59
    πει
    0.58
    lero
    0.57
     cuore
    0.55
    Activations Density 0.319%

    No Known Activations