© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 1299
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on numeric tokens and location or proper‐name tokens—i.e. references to specific places, dates, measurements, and quantities.

    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
    TARGETTING
    0.80
     воздей
    0.68
     Україні
    0.66
     ⿱
    0.65
    INFE
    0.64
     최종
    0.63
     要素
    0.63
     সংগ্রামে
    0.62
     攵
    0.62
     द्विवेदी
    0.61
    POSITIVE LOGITS
     cemetery
    1.65
     Museum
    1.61
     museum
    1.59
     airport
    1.55
     promenade
    1.54
     museums
    1.52
     cemeteries
    1.48
     waterfront
    1.48
     Cemetery
    1.46
     park
    1.42
    Activations Density 1.235%

    No Known Activations