© 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. 2847
    Prev
    Next
    INDEX
    Explanations

    finish or devour

    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
    æĬ½
    -0.10
    Carol
    -0.09
    æķ´åIJĪ
    -0.09
    å·¡
    -0.09
    /bus
    -0.08
    ãĤ¢ãĥĬ
    -0.08
     Explorer
    -0.08
    åĮ¾
    -0.08
    å±İ
    -0.08
    è£Ļ
    -0.08
    POSITIVE LOGITS
    ä¸Ģåı£æ°Ķ
    0.14
    ä¸Ģåı£
    0.12
    åIJĥå®Į
    0.11
    åIJŀ
    0.10
    finish
    0.10
     Ere
    0.10
     finishing
    0.09
    åIJĥå¾Ĺ
    0.09
    ucid
    0.09
    ä¸ĢæĿ¯
    0.09
    Activations Density 0.057%

    No Known Activations