© 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. 1698
    Prev
    Next
    INDEX
    Explanations

    complex tasks and difficulties

    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
    å¾Īéĩįè¦ģçļĦ
    -0.09
    æĺ¯æľĢ好çļĦ
    -0.09
    èĿĩ
    -0.09
    æĬ¬å¤´
    -0.08
    éĦĦ
    -0.08
    毫ä¸į
    -0.08
    ä»Ģ
    -0.08
    æĹłéľĢ
    -0.08
    转载请
    -0.08
    ç»Łè®¡æķ°æį®
    -0.08
    POSITIVE LOGITS
    èĢĥéªĮ
    0.13
     пÑĢидеÑĤÑģÑı
    0.11
    ä¸į容æĺĵ
    0.11
    éļ¾åº¦
    0.11
     required
    0.10
     requires
    0.10
    åĽ°éļ¾
    0.10
    æĹ¢è¦ģ
    0.10
    difficulty
    0.10
    èī°å·¨
    0.10
    Activations Density 0.287%

    No Known Activations