© 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. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 20471
    Prev
    Next
    INDEX
    Explanations

    referee, judge, manager roles

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    é²²
    -0.10
    liest
    -0.10
    биÑĤ
    -0.10
    orning
    -0.09
    æĦı大
    -0.09
    hiba
    -0.09
     игÑĢÑĭ
    -0.09
    inkel
    -0.09
    æľīæĿ¡ä»¶çļĦ
    -0.09
    ictions
    -0.09
    POSITIVE LOGITS
    们
    0.10
    rey
    0.10
    weed
    0.10
     Misc
    0.08
    ersed
    0.08
    eters
    0.08
    ãģĵãģ¨
    0.08
    è§ŀ
    0.08
    pon
    0.08
    âĸº
    0.08
    Activations Density 0.001%

    No Known Activations