© 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. 2149
    Prev
    Next
    INDEX
    Explanations

    cl followed by clusters, clamp, or clustering

    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
    ium
    -0.10
    æĹĥ
    -0.09
    едÑĮ
    -0.09
     elk
    -0.08
    èĨı
    -0.08
    å°ıåºĹ
    -0.08
    ä¸Ģå¾ĭ
    -0.07
    éĤ¨
    -0.07
    \AppData
    -0.07
    /close
    -0.07
    POSITIVE LOGITS
    usters
    0.11
    è¡Įä¸ļåįıä¼ļ
    0.10
    /cl
    0.10
    USTER
    0.10
    rooms
    0.10
    rok
    0.10
    ervo
    0.10
    ave
    0.10
    iffs
    0.09
    buster
    0.09
    Activations Density 0.061%

    No Known Activations