© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 20233
    Prev
    Next
    INDEX
    Explanations

    will, can, would, or 's followed by verb

    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
    edin
    -0.09
    èĭ¥æĤ¨
    -0.09
    _cuda
    -0.09
    lu
    -0.09
     Fur
    -0.08
    让æĤ¨
    -0.08
    ä¸įä¸ĭ
    -0.08
    .float
    -0.08
    _GRAPH
    -0.08
    /locale
    -0.08
    POSITIVE LOGITS
    éĻª
    0.11
    大éĺŁ
    0.11
    æ½ľ
    0.10
    æ³¥
    0.09
    ä¸ĢåıĮ
    0.09
    æļĹæļĹ
    0.09
    éĶ»
    0.09
    æĹłè¾ľ
    0.09
    ayne
    0.09
    å¬ĸ
    0.09
    Activations Density 0.024%

    No Known Activations