© 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. 23519
    Prev
    Next
    INDEX
    Explanations

    he they introduce actions

    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
    imin
    -0.10
    etter
    -0.10
    æĺ¯ä¸Ģ件
    -0.10
    è·ŁçĿĢ
    -0.10
     erfol
    -0.10
    à¹Ħล
    -0.09
    建ç«ĭèµ·
    -0.09
    éĥ¢
    -0.09
    ledo
    -0.09
     Ngo
    -0.08
    POSITIVE LOGITS
    ä¸į说
    0.10
     recap
    0.09
    çľĭäºĨä¸Ģ
    0.09
    çľĭå¾Ĺ
    0.09
    带åĪ°
    0.09
    缮
    0.09
    ä¸įå®Į
    0.09
    urple
    0.09
    irs
    0.09
    iku
    0.09
    Activations Density 0.115%

    No Known Activations