© 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. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 56954
    Prev
    Next
    INDEX
    Explanations

    in or go or back or hit

    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
    ÂŃs
    -0.10
    ickle
    -0.09
    umer
    -0.09
    æĺ¯åIJ¦ä¼ļ
    -0.09
    ãĥ¼ãĥł
    -0.08
    	constructor
    -0.08
    ourd
    -0.08
    icted
    -0.08
    edly
    -0.08
    (Constructor
    -0.08
    POSITIVE LOGITS
    ä¹ĭæŶ
    0.10
    è¿ĽæĿ¥
    0.10
    æĢ§å¼º
    0.10
    ä¸įæĥľ
    0.09
    èĢĮåİ»
    0.09
    è¿Ľåİ»
    0.09
    éĢ¡
    0.09
    ä¸İåIJ¦
    0.09
    ä¸įç¦ģ
    0.09
    åİ»äºĨ
    0.09
    Activations Density 0.112%

    No Known Activations