© 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. 57466
    Prev
    Next
    INDEX
    Explanations

    hone, refine, sharpen

    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
    çļĦåıĺåĮĸ
    -0.09
    æ¬
    -0.09
    ity
    -0.08
    ạng
    -0.08
    ahr
    -0.08
    åĴĮæĬĢæľ¯
    -0.08
    roy
    -0.08
    atem
    -0.08
    ä½łæĢİä¹Ī
    -0.08
    å¿ĥåĬ¨
    -0.08
    POSITIVE LOGITS
    åĴĮå®ĮåĸĦ
    0.15
    å®ĮåĸĦ
    0.14
    å®ĮåĸĦçļĦ
    0.13
     skills
    0.11
    chter
    0.11
    æĬĢèīº
    0.11
     refine
    0.10
     Skills
    0.10
    çłº
    0.10
    å¯Ĥå¯ŀ
    0.10
    Activations Density 0.013%

    No Known Activations