© 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. 30564
    Prev
    Next
    INDEX
    Explanations

    out of character or departure

    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
    æ¯Ķäºļ
    -0.11
    主è§Ĥ
    -0.10
    _DRIVE
    -0.10
    å¾Ĺ天
    -0.09
    orate
    -0.09
    nameof
    -0.09
    æĮĤåľ¨
    -0.09
    ä¸įæĸŃå®ĮåĸĦ
    -0.09
    饼
    -0.09
    VIC
    -0.09
    POSITIVE LOGITS
    ç½ķè§ģ
    0.15
    å°ijè§ģ
    0.14
     normally
    0.13
    å¹³æŶ
    0.13
    normally
    0.12
    åīįæīĢæľª
    0.12
    åīįæīĢæľªæľī
    0.12
    ç½ķ
    0.12
     Normally
    0.11
    éĻĮçĶŁ
    0.11
    Activations Density 0.094%

    No Known Activations