© 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. 3489
    Prev
    Next
    INDEX
    Explanations

    out 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
    骤
    -0.10
    åĿł
    -0.09
    第äºĮ次
    -0.09
    ãģĸ
    -0.09
    æķŀ
    -0.09
    åĿŁ
    -0.08
    åħ¶éĹ´
    -0.08
    åĴĮå®ĮåĸĦ
    -0.08
    è§ĤçĤ¹
    -0.08
    ëĦ¤
    -0.08
    POSITIVE LOGITS
    举缣
    0.11
    éļ³
    0.09
    OOK
    0.09
    ç¥ŀ
    0.09
    ä¸įéĶĻ
    0.09
    té
    0.09
    çīĮ
    0.08
    aoke
    0.08
    pute
    0.08
    essional
    0.08
    Activations Density 0.008%

    No Known Activations