© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 36129
    Prev
    Next
    INDEX
    Explanations

    contrast introducing limitations

    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
    èĢĮè¿Ļ
    -0.10
    è¿Ļæł·ä¸Ģ
    -0.10
    olean
    -0.10
    ç
    -0.09
    ä¹ŁåĽłæѤ
    -0.09
    lates
    -0.09
    ä¿Ŀ管
    -0.09
    ÙĦاÙģ
    -0.09
    olian
    -0.09
    ĥģ
    -0.09
    POSITIVE LOGITS
    ä¸įä¸Ģå®ļ
    0.10
    ä¸įåıĬ
    0.09
    ä¸įæİĴéĻ¤
    0.09
    æį®è¯´
    0.08
    /or
    0.08
    ä¸įä¸Ĭ
    0.08
    BST
    0.08
     mostly
    0.07
    çīĪæĿĥ声æĺİ
    0.07
    æĽ´å¤ļçļĦæĺ¯
    0.07
    Activations Density 0.116%

    No Known Activations