© 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. 27139
    Prev
    Next
    INDEX
    Explanations

    data-based, evidence-based, objective measurement

    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
    ImagePath
    -0.10
    èĤ¢
    -0.10
    æµħ
    -0.09
    çļĦæĸ°
    -0.09
    åıĹ访èĢħ
    -0.09
    顺åĪ©
    -0.09
    æĪĸå¤ļ
    -0.09
    ä¹IJæĦı
    -0.09
    娥
    -0.09
    离éĢĢä¼ij
    -0.09
    POSITIVE LOGITS
    ç§ijåѦ
    0.18
     scientific
    0.14
     science
    0.13
    ç§ijåѸ
    0.13
    ç§ijåѦçłĶ究
    0.12
    çIJĨæĢ§
    0.12
    循
    0.12
     Scientific
    0.12
     facts
    0.12
    äºĭå®ŀ
    0.12
    Activations Density 0.118%

    No Known Activations