© 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. 24099
    Prev
    Next
    INDEX
    Explanations

    complicity, aiding, illicit, acquiesce

    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
    ç¹ģ
    -0.08
    äºĶè¡Į
    -0.08
    åĹ£
    -0.08
     ausp
    -0.08
     Kr
    -0.08
    æĸĭ
    -0.08
    æĭ¥æľī
    -0.08
    æĿ¥ä¸´
    -0.08
    楼
    -0.08
    riad
    -0.08
    POSITIVE LOGITS
    åĬ©éķ¿
    0.17
    容å¿į
    0.13
    åħģ许
    0.12
    .accept
    0.12
    çĤ¹å¤´
    0.12
    é»ĺå¥ij
    0.12
     accept
    0.11
    æݥ纳
    0.11
    容纳
    0.11
     позвол
    0.11
    Activations Density 0.092%

    No Known Activations