© 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. 12134
    Prev
    Next
    INDEX
    Explanations

    found when surprised

    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
    çݺ
    -0.10
    æݾ
    -0.09
    aris
    -0.09
     zwar
    -0.09
    éķ§
    -0.08
    èĤī
    -0.08
    Camb
    -0.08
    ç¾Ŀ
    -0.08
    áŀĦ
    -0.08
     ëıĦ
    -0.08
    POSITIVE LOGITS
    å½ĵæĪij
    0.13
    when
    0.10
    å½ĵä»ĸ
    0.09
     when
    0.09
    hana
    0.09
    å½ĵæĪij们
    0.09
    è¿İæĿ¥äºĨ
    0.09
    çIJĨä¼ļ
    0.08
    ä¼ļ对
    0.08
    èģļåIJĪ
    0.08
    Activations Density 0.073%

    No Known Activations