© 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. 49284
    Prev
    Next
    INDEX
    Explanations

    licensed to

    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
    è¿ĻéĩĮ
    -0.17
    çİ°åľ¨
    -0.13
    æĿ¥èĩª
    -0.11
    æľĢåIJİä¸Ģ个
    -0.11
    alyze
    -0.10
    åıĸãĤĬ
    -0.10
    inate
    -0.10
    ãģĶ
    -0.10
    ãģ¤ãģij
    -0.10
    åı¯ä»¥è¯´
    -0.09
    POSITIVE LOGITS
    /from
    0.27
    gether
    0.22
    plevel
    0.18
    åģļçļĦäºĭ
    0.16
    缮åīį为æŃ¢
    0.15
    iling
    0.13
    åģļçļĦäºĭæĥħ
    0.13
    asting
    0.13
    -do
    0.13
    xic
    0.13
    Activations Density 1.680%

    No Known Activations