© 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. 54956
    Prev
    Next
    INDEX
    Explanations

    getting it right or wrong

    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
    æ½ľèĥ½
    -0.10
     Bott
    -0.09
    æªł
    -0.09
    çĵ¶é¢Ī
    -0.09
    ç¼¥
    -0.09
     sprung
    -0.09
    æīĵ
    -0.09
     chai
    -0.08
    åѦåłĤ
    -0.08
    æĽ¹
    -0.08
    POSITIVE LOGITS
    ä¸į对
    0.11
    éĶĻäºĨ
    0.10
     matters
    0.10
    è¹ij
    0.09
    äºĭ项
    0.09
    ä¸ĢåĽŀäºĭ
    0.09
    æĦı大
    0.09
     wrong
    0.09
    没éĶĻ
    0.08
    åģ¾
    0.08
    Activations Density 0.044%

    No Known Activations