© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 21882
    Prev
    Next
    INDEX
    Explanations

    due to, contrast to, keeping a

    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
    æ¯Ķå¦Ĥ
    -0.09
    æ¯Ķå¦Ĥ说
    -0.08
    çļĦæĸ¹å¼ıæĿ¥
    -0.08
    TRL
    -0.08
    èѬå¦Ĥ
    -0.08
    æ·¡å®ļ
    -0.08
    æĦł
    -0.08
    幡
    -0.08
     flagged
    -0.08
    就好
    -0.08
    POSITIVE LOGITS
    ä»Ģä¹ĪåİŁåĽł
    0.11
    çĶ±äºİ
    0.11
     due
    0.10
    éī´äºİ
    0.10
     Due
    0.10
    Due
    0.10
    æĹłå¥Ī
    0.09
    è¿İåIJĪ
    0.09
    éĺī
    0.09
    为äºĨéģ¿åħį
    0.09
    Activations Density 0.204%

    No Known Activations