© 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. 63370
    Prev
    Next
    INDEX
    Explanations

    casting spells or making eye contact

    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
    æŁIJç§įç¨ĭ度
    -0.10
    ä¸Ģç¬ij
    -0.10
    æŁIJ个
    -0.09
    вÑĪ
    -0.09
    éĤĭ
    -0.09
    .mods
    -0.09
    æī¾ä¸ª
    -0.09
    è¿Ļåĩłä¸ª
    -0.09
    ãģ¨æĢĿãģ£ãģŁ
    -0.09
    è¾ĥå¤ļ
    -0.09
    POSITIVE LOGITS
    åIJĦç§į
    0.12
    ä¸Ģæĸ¹éĿ¢
    0.12
    ä¸įæĸŃåľ°
    0.12
    ä¸įåIJĮçļĦ
    0.11
    ä¸Ģ次次
    0.10
    ä¸įåģľåľ°
    0.10
    ä¸įæĸŃ
    0.10
    ãģ©ãĤĵãģ©ãĤĵ
    0.10
    è¶Ĭ
    0.10
    åIJĦæł·
    0.10
    Activations Density 0.088%

    No Known Activations