© 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. 33363
    Prev
    Next
    INDEX
    Explanations

    empire

    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
    åİŁåĪĻä¸Ĭ
    -0.10
    sure
    -0.09
    à¸Ļาà¸Ļ
    -0.09
     Rings
    -0.08
    su
    -0.08
    yc
    -0.08
    -than
    -0.08
     trÃŃ
    -0.08
    á»Ŀ
    -0.08
    çĿ¬
    -0.08
    POSITIVE LOGITS
    -wide
    0.11
    strate
    0.10
    wide
    0.10
    éľīç´ł
    0.09
    æĿī
    0.09
     minded
    0.09
    客æĪ·æıIJä¾Ľ
    0.09
    -building
    0.09
    åĮĸçļĦ
    0.09
    Lite
    0.09
    Activations Density 0.006%

    No Known Activations