© 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. 20591
    Prev
    Next
    INDEX
    Explanations

    descriptors like 'ugly' or 'pretty'

    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
    ä¸įå¦Ĥ
    -0.10
     drib
    -0.09
    æł¹æľ¬ä¸į
    -0.09
    ä¸įä»ħ
    -0.08
     Bott
    -0.08
    ä¸įä½Ĩ
    -0.08
    æĹ¢æľī
    -0.08
     naturally
    -0.08
    Twig
    -0.08
    è¿IJ
    -0.08
    POSITIVE LOGITS
    æ¯ķç«Łæĺ¯
    0.15
    æ¯ķç«Ł
    0.12
    åį´æĺ¯
    0.10
    æľīä»·åĢ¼çļĦ
    0.10
    çķ¢ç«Ł
    0.10
     nonetheless
    0.10
    ä»įå°Ĩ
    0.10
     nevertheless
    0.10
    æľīä»·åĢ¼
    0.10
    çķ¢
    0.10
    Activations Density 0.088%

    No Known Activations