© 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. 54490
    Prev
    Next
    INDEX
    Explanations

    Madden, Outlook, Deadpool, Citadel

    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
    ç¼İ
    -0.13
    羣æŃ£åģļåĪ°
    -0.09
    Ñģен
    -0.09
    ìļ¸
    -0.09
    ERING
    -0.09
    æĢ»ä¹ĭ
    -0.08
    æĿĵ
    -0.08
    ueil
    -0.08
    æĸ°åĮº
    -0.08
    ussen
    -0.08
    POSITIVE LOGITS
    ../../../
    0.11
    elan
    0.09
    ราว
    0.09
    趣åij³
    0.08
    yme
    0.08
    phis
    0.08
    ä¸įåIJĮç¨ĭ度
    0.08
    chy
    0.08
    趣
    0.08
    åķī
    0.08
    Activations Density 0.009%

    No Known Activations