© 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. 55250
    Prev
    Next
    INDEX
    Explanations

    bare-knuckled boxer

    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
    ä½ĵè´¨
    -0.09
    æĢ»çļĦ
    -0.09
    icks
    -0.08
    çIJı
    -0.08
    poster
    -0.08
    è´¨
    -0.08
     Roberto
    -0.08
    èĺ§
    -0.08
     pocket
    -0.08
    obody
    -0.08
    POSITIVE LOGITS
    æĹłåĬ©
    0.09
    çĬ¶æĢģä¸ĭ
    0.09
    奢
    0.09
    ï¸
    0.09
    ä¸ĢæĹı
    0.09
    åľ¨æ·±åľ³
    0.09
    浪费
    0.09
    à¸Ĺะà¹Ģล
    0.09
    ÑĤÑĥ
    0.09
    åĶ®åIJİ
    0.09
    Activations Density 0.007%

    No Known Activations