© 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. 5757
    Prev
    Next
    INDEX
    Explanations

    certain list items like : or *

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

    The neuron is highly activated by numeric tokens or numerical references (e.g., years, page numbers, percentages, citation numbers) 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
    부터
    1.10
    liği
    0.84
    دي
    0.83
    デア
    0.83
     劉
    0.81
    VSLU
    0.81
    <0xB4>
    0.80
    テレ
    0.80
     Partly
    0.80
    диа
    0.79
    POSITIVE LOGITS
    signature
    0.87
    password
    0.73
    heading
    0.73
    kwargs
    0.73
    le
    0.72
    idable
    0.72
     nitrogen
    0.71
    ppings
    0.71
    soluble
    0.70
    u
    0.70
    Activations Density 0.001%

    No Known Activations