© 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. 5066
    Prev
    Next
    INDEX
    Explanations

    over-expression, overcharge, overtrained, overworked, overnutrition

    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
    æĽ´å¤ļ
    -0.12
    大åĬĽæĶ¯æĮģ
    -0.09
    ubi
    -0.09
     Hers
    -0.09
    asurement
    -0.08
     dances
    -0.08
    主ä¹īæĢĿæĥ³
    -0.08
    æĶ¸
    -0.08
    dance
    -0.08
    ä¸įå¼Ģ
    -0.08
    POSITIVE LOGITS
    ening
    0.09
    ingly
    0.09
    以èĩ³äºİ
    0.08
     tÆ°á»Łng
    0.08
    马路
    0.08
    ste
    0.08
     Nice
    0.08
    nice
    0.08
    èĵĿèī²
    0.08
    çĶļèĩ³
    0.08
    Activations Density 0.061%

    No Known Activations