© 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. 29467
    Prev
    Next
    INDEX
    Explanations

    proper way or correctly

    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
     ÐĺÑģ
    -0.09
     Came
    -0.09
    _chance
    -0.09
    éĤ£
    -0.08
    çļĦçī¹çĤ¹
    -0.08
    è°İ
    -0.08
    wm
    -0.08
    ä¸ĢéĹª
    -0.08
    hex
    -0.08
    å¹³éĿĻ
    -0.08
    POSITIVE LOGITS
    æŃ£ç¡®çļĦ
    0.17
     correctly
    0.15
     proper
    0.15
    æŃ£ç¡®
    0.15
    proper
    0.15
     пÑĢавилÑĮно
    0.14
     properly
    0.14
    æŃ£ç¢º
    0.14
     Proper
    0.14
    ä¸įå½ĵ
    0.13
    Activations Density 0.100%

    No Known Activations