© 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. 40161
    Prev
    Next
    INDEX
    Explanations

    guarantee you

    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
    ç½ij讯
    -0.11
    åįĥéĩĮ
    -0.09
    ERA
    -0.09
    浪
    -0.09
    家ä¼Ļ
    -0.08
    åIJ¬è¯Ŀ
    -0.08
    çļĦéĴ±
    -0.08
    æįŁå¤±
    -0.08
    çļĦéĤ£ä¸Ģ
    -0.08
    尾巴
    -0.08
    POSITIVE LOGITS
    indr
    0.08
    gang
    0.08
    å¸
    0.08
    impse
    0.08
    rito
    0.08
    ogui
    0.08
    iced
    0.08
    iled
    0.08
    èĨĿ
    0.08
    _ini
    0.08
    Activations Density 0.006%

    No Known Activations