© 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. 54982
    Prev
    Next
    INDEX
    Explanations

    exceeds expectations

    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
    æIJŃ
    -0.09
    æİĪ
    -0.08
    omic
    -0.08
    æľīæĦıæĢĿ
    -0.08
    åį¿
    -0.08
    å¿ħä¸įåı¯
    -0.08
    æľīåIJį
    -0.08
    éĻª
    -0.08
    å®Ĺ
    -0.08
    ä¼ĺåĬ¿
    -0.08
    POSITIVE LOGITS
    é¢ĦæľŁ
    0.18
     expectations
    0.16
     expectation
    0.12
    unexpected
    0.12
    é¢Ħå®ļ
    0.12
    è¶ħåĩº
    0.11
     unexpected
    0.11
     unexpectedly
    0.11
    é«ĺäºİ
    0.11
    å¦ĤæľŁ
    0.11
    Activations Density 0.187%

    No Known Activations