© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 52279
    Prev
    Next
    INDEX
    Explanations

    top key most staple

    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
    pecia
    -0.10
    ufs
    -0.10
    ancia
    -0.09
    edom
    -0.09
    ymb
    -0.09
    arta
    -0.09
    è¾ĸ
    -0.08
    满æĢĢ
    -0.08
     dzie
    -0.08
    .hw
    -0.08
    POSITIVE LOGITS
    å¿ħä¸įåı¯å°ij
    0.10
    稳å®ļçļĦ
    0.10
    ä¸įåı¯ç¼ºå°ij
    0.10
    å½ĵä¹ĭæĹłæĦ§
    0.09
    å¿ħä¸įåı¯
    0.09
    æĮĩæľĽ
    0.09
    åĴĮåıijå±ķ
    0.09
    æŀķ
    0.09
    å¼ķé¢Ĩ
    0.09
    ä¸İåıijå±ķ
    0.09
    Activations Density 0.025%

    No Known Activations