© 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. 14248
    Prev
    Next
    INDEX
    Explanations

    job location

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

    The neuron is specialized to pick out numeric data tokens—especially measurements, counts, or numeric values (including decimals and fractions)—in the text.

    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
    allenges
    0.88
    ifically
    0.74
    imente
    0.74
    ǘ
    0.73
    ipers
    0.71
    ötz
    0.71
    agation
    0.70
    odiac
    0.70
     ഹോളി
    0.70
    imbledon
    0.69
    POSITIVE LOGITS
     both
    0.88
     neither
    0.86
     både
    0.66
     zarówno
    0.65
     latter
    0.62
     help
    0.61
    ↵↵
    0.59
    <0xC2>
    0.59
     external
    0.59
     first
    0.56
    Activations Density 0.197%

    No Known Activations