© 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. 724
    Prev
    Next
    INDEX
    Explanations

    now followed by current name

    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
     formerly
    -0.11
    æľ¬æĿ¥
    -0.10
    åİŁæľ¬
    -0.10
    ä¸Ģ度
    -0.10
     originally
    -0.10
    éĥIJ
    -0.09
    arcy
    -0.09
    former
    -0.09
    æĽ¾ç¶ĵ
    -0.09
    åİŁ
    -0.08
    POSITIVE LOGITS
    æ®Ĭ
    0.11
    æĥĺ
    0.09
    èĮĤ
    0.09
    arch
    0.09
    æį¢æĪIJ
    0.09
     авÑĤом
    0.08
    TF
    0.08
    æĶ¾æīĭ
    0.08
    çļĦéĩįçĤ¹
    0.08
    adi
    0.08
    Activations Density 0.065%

    No Known Activations