© 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. 63281
    Prev
    Next
    INDEX
    Explanations

    inaccuracy and incorrectness

    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
    驼
    -0.10
    羣å¿ĥ
    -0.09
    èĢĮåIJİ
    -0.08
    殳
    -0.08
    羣çļĦå¾Ī
    -0.08
    æľµ
    -0.08
    legg
    -0.08
    翦
    -0.08
    â»
    -0.08
    æĶ¶åĪ°äºĨ
    -0.08
    POSITIVE LOGITS
    æ¦Ĥ
    0.09
    Ñģов
    0.09
    议论
    0.09
    èĨĤ
    0.09
    ä¸ĸ人
    0.09
     ÐĴлад
    0.09
    ä¼ļåIJĮ
    0.09
    ÙĪØŃ
    0.09
    cription
    0.08
    æ¯ħçĦ¶
    0.08
    Activations Density 0.004%

    No Known Activations