© 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. 55837
    Prev
    Next
    INDEX
    Explanations

    earth, forces, layer, piece

    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
    é¦
    -0.08
    loy
    -0.07
    -0.07
    è¿Ļä¸ĢçĤ¹
    -0.06
     >>=
    -0.06
    3
    -0.06
    ê·Ģ
    -0.06
    1
    -0.06
     bail
    -0.06
    åĨ
    -0.06
    POSITIVE LOGITS
    EOS
    0.10
    ä¸Ńæĸ°ç½ij
    0.10
    ãģĶè¦
    0.09
    ãĥĵãĤ¸ãĥį
    0.09
    toolbox
    0.09
    æĥĬåı¹
    0.09
     ìĿ´ìķ¼
    0.09
    ãįij
    0.09
    å¨ĵ
    0.09
    ä¸ĢèĪ¬éĥ½æĺ¯
    0.09
    Activations Density 1.096%

    No Known Activations