© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 6648
    Prev
    Next
    INDEX
    Explanations

    attracted to specific groups

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

    The neuron primarily activates on numeric tokens—especially decimal numbers or figures used as data points, section references, or measurements.

    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
    Об
    0.75
    𝙄
    0.75
     যথা
    0.73
    要把
    0.73
     коллеги
    0.71
    Го
    0.71
     pere
    0.71
    变换
    0.71
     spiced
    0.70
    𝙝
    0.70
    POSITIVE LOGITS
    n
    0.69
     isomers
    0.66
    ตำแหน่ง
    0.66
    นะนำ
    0.66
    DURATION
    0.66
    ます
    0.65
     Akan
    0.65
    ている
    0.63
    主要
    0.63
     Deze
    0.63
    Activations Density 0.001%

    No Known Activations