© 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. 4208
    Prev
    Next
    INDEX
    Explanations

    guilty or violating rules

    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
    çļĦåIJĦ项
    -0.11
     اÙĦØ£ÙĪ
    -0.10
    èĿł
    -0.10
    æīĢéľĢè¦ģçļĦ
    -0.09
    èIJ½åľ°
    -0.09
    çļĦ人çĶŁ
    -0.09
    åķĨæľº
    -0.09
    åIJĦæł·
    -0.09
    ä»İ容
    -0.09
    æĨ¶
    -0.09
    POSITIVE LOGITS
    åľ¨è¿Ļæĸ¹éĿ¢
    0.14
    è¿Ļæĸ¹éĿ¢
    0.13
    çĬ¯
    0.12
     thói
    0.12
    æ¯ĽçĹħ
    0.12
     habit
    0.11
     guilty
    0.11
    表çİ°
    0.10
    è¿Ļä¸ĢçĤ¹
    0.10
     tendency
    0.10
    Activations Density 0.081%

    No Known Activations