© 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. 28080
    Prev
    Next
    INDEX
    Explanations

    first time/question/report

    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
    æľĢçµĤ
    -0.10
    第ä¸īå±Ĭ
    -0.10
    ä¸İåIJ¦
    -0.09
    é¦ĸå±Ĭ
    -0.09
    æľĢç»Ī
    -0.09
    éĺ¡
    -0.09
    ÑĤик
    -0.09
     Rai
    -0.09
    第ä¸ĥå±Ĭ
    -0.09
    åIJ¸
    -0.09
    POSITIVE LOGITS
    -ever
    0.15
     ever
    0.14
     tiên
    0.13
    Ever
    0.11
     EVER
    0.11
     Ever
    0.10
    -rate
    0.10
     glance
    0.09
    ever
    0.09
     baseman
    0.09
    Activations Density 0.449%

    No Known Activations