© 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. 22408
    Prev
    Next
    INDEX
    Explanations

    offense/offensive play

    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
    æĺ¶
    -0.12
    atan
    -0.10
    åĵģæł¼
    -0.09
    ä¹ĭ乡
    -0.08
    dio
    -0.08
    äºĨ好åĩł
    -0.08
    Rated
    -0.08
    é¾ĻéŨ
    -0.08
    é¾Ļæ¹ĸ
    -0.08
    orary
    -0.08
    POSITIVE LOGITS
    æ·±å±Ĥ次
    0.11
    uell
    0.10
    èĸª
    0.09
     ÑģовеÑĢÑĪен
    0.09
    .ut
    0.09
    æĸ¹éĿ¢
    0.09
    åij½
    0.08
    ional
    0.08
    æݾ
    0.08
    ç¹ģ
    0.08
    Activations Density 0.007%

    No Known Activations