© 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. 3678
    Prev
    Next
    INDEX
    Explanations

    as an example of a general rule

    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
    ä¹ĥ
    -0.09
    ITO
    -0.09
    GMEM
    -0.09
    LOT
    -0.09
     MATERIAL
    -0.09
    енÑĮк
    -0.08
    èĮĨ
    -0.08
    conc
    -0.08
    #g
    -0.08
    .live
    -0.08
    POSITIVE LOGITS
    家åĬ¡
    0.10
    ä¸ī家
    0.10
    ramer
    0.09
    åıªèĥ½è¯´
    0.09
    æĸ¹éĴĪ
    0.09
     sanity
    0.08
    åºķèķ´
    0.08
    纡
    0.08
    éļ¶
    0.08
    æĿ¥è®²
    0.08
    Activations Density 0.042%

    No Known Activations