© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 26200
    Prev
    Next
    INDEX
    Explanations

    loss function, backpropagation, activations

    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
    Negative Logits
    ç½Ĺ
    -0.10
    åĽ¾æ¡Ī
    -0.10
    èĢ³æľµ
    -0.10
    orda
    -0.09
    é¾Ľ
    -0.09
    éĥ¤
    -0.08
    èĢ³
    -0.08
    vens
    -0.08
    æīİ
    -0.08
    ocache
    -0.08
    POSITIVE LOGITS
    æĶ¶åıĸ
    0.09
    çī¹åĪ«å£°æĺİ
    0.08
    æĶ¸
    0.08
    贯穿
    0.08
    å±ĤéĿ¢
    0.08
    aji
    0.08
    å»ī
    0.08
     Eg
    0.08
    ä»Ķ
    0.08
    设计çIJĨ念
    0.08
    Activations Density 0.110%

    No Known Activations