© 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. 20017
    Prev
    Next
    INDEX
    Explanations

    reprint or print

    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
    æĺ¯åIJ¦åŃĺåľ¨
    -0.09
    èĹı
    -0.09
    éĩĩ访
    -0.09
    ç¼ļ
    -0.08
    itian
    -0.08
    Defs
    -0.08
    itor
    -0.08
    èĸª
    -0.08
    irk
    -0.08
     Braun
    -0.08
    POSITIVE LOGITS
    aro
    0.09
    éré
    0.09
     thr
    0.09
    鸩
    0.09
     Naomi
    0.09
    aire
    0.09
    åĪ¨
    0.08
    马æĭīæĿ¾
    0.08
    rij
    0.08
    Kernel
    0.08
    Activations Density 0.031%

    No Known Activations