© 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. 58835
    Prev
    Next
    INDEX
    Explanations

    omission, slope, marginal, plenty

    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
    æ¡Ī
    -0.09
    çļĦå®ŀéĻħ
    -0.09
     Та
    -0.09
    _trampoline
    -0.08
    orious
    -0.08
    åĨ³
    -0.08
    çªŁ
    -0.08
    é¢ĨæĤŁ
    -0.08
    _metric
    -0.08
    çļĦéĤ£ä¸ª
    -0.08
    POSITIVE LOGITS
    äºĮèĢħ
    0.09
    ause
    0.09
    ä¸ŃåĽ½ç»ıæµİ
    0.09
    upy
    0.09
    æĺ¥èĬĤæľŁéĹ´
    0.09
    .gc
    0.09
    ç§ijæĬĢåĽŃ
    0.09
    çĬĬ
    0.08
    ä»ĵä½į
    0.08
    è±¹
    0.08
    Activations Density 0.146%

    No Known Activations