© 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. 57618
    Prev
    Next
    INDEX
    Explanations

    ever / highest / first

    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
    ến
    -0.10
    دÙĬÙĨ
    -0.09
     tav
    -0.09
    ä¼ļä¸įä¼ļ
    -0.08
    lam
    -0.08
    Drv
    -0.08
    å¦ĤæŀľæĤ¨
    -0.08
     вÑĭпÑĥÑģ
    -0.08
    åij¨æľŁ
    -0.08
    hee
    -0.08
    POSITIVE LOGITS
    åIJĮç±»
    0.10
    åİĨæĿ¥
    0.10
    åľ¨åIJĮä¸Ģ
    0.10
    åIJĮä¸Ģ个
    0.09
     anywhere
    0.09
    åľ¨ä¸Ģ个
    0.09
     ever
    0.09
    æŁĶ
    0.09
    åIJĮä¸Ģ
    0.09
     EVER
    0.08
    Activations Density 0.194%

    No Known Activations