© 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. 18602
    Prev
    Next
    INDEX
    Explanations

    as a result of

    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
    ature
    -0.10
    trail
    -0.09
    ãģĹãģ¾ãģĨ
    -0.09
    rice
    -0.09
    ç»ĵæŀľæĺ¾ç¤º
    -0.09
    çĤ³
    -0.09
    æĸ¹éĿ¢çļĦ
    -0.09
     reel
    -0.09
    ilen
    -0.09
    绦
    -0.09
    POSITIVE LOGITS
    ively
    0.12
     Ñĩего
    0.11
    noon
    0.10
    ion
    0.09
    fully
    0.09
    leigh
    0.09
    ive
    0.09
     either
    0.09
    inium
    0.09
    ful
    0.08
    Activations Density 0.054%

    No Known Activations