© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 30610
    Prev
    Next
    INDEX
    Explanations

    Korean or joo/jun names

    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
    çģĮæºī
    -0.09
    åłķ
    -0.09
     DAL
    -0.09
    ëĭī
    -0.08
    æĽ´ä½ķåĨµ
    -0.08
    ermann
    -0.08
    odox
    -0.08
    fell
    -0.08
    ers
    -0.08
    æĬ¬å¤´
    -0.08
    POSITIVE LOGITS
    æĺ¯éŁ©åĽ½
    0.20
    éŁ©åĽ½
    0.15
    飩
    0.13
     HÃłn
    0.12
    æľĿé²ľ
    0.12
     Korea
    0.12
     Seoul
    0.12
     Korean
    0.12
    éŁĵ
    0.11
     Kim
    0.11
    Activations Density 0.076%

    No Known Activations