© 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. 745
    Prev
    Next
    INDEX
    Explanations

    loyal, faithful, devoted

    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
    çģµæĦŁ
    -0.10
    éĹ°
    -0.09
    æĴ¤
    -0.09
    ä¼ļåIJĮ
    -0.09
     Wan
    -0.09
    çݯå¢ĥ
    -0.08
    iff
    -0.08
     sco
    -0.08
    夺
    -0.08
    åĨķ
    -0.08
    POSITIVE LOGITS
    å¿łè¯ļ
    0.27
     loyalty
    0.22
    å¿ł
    0.20
    loy
    0.19
    å¿łå®ŀ
    0.18
     loyal
    0.18
     allegiance
    0.15
    ç»´æĬ¤
    0.14
    LOY
    0.14
     Loy
    0.14
    Activations Density 0.044%

    No Known Activations