© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 59086
    Prev
    Next
    INDEX
    Explanations

    was/were/been followed by adjective/adverb

    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
     nghá»ģ
    -0.09
    é¾Ħ
    -0.09
     collapsing
    -0.09
    ofday
    -0.09
    åķĨ
    -0.08
    alem
    -0.08
    代è¨Ģ
    -0.08
    æĸ¹å½¢
    -0.08
    èģĮä¸ļéģĵå¾·
    -0.08
    ено
    -0.08
    POSITIVE LOGITS
    ä¸Ģå¥Ĺ
    0.09
     able
    0.08
    ä¸Ģåı°
    0.08
    大纲
    0.08
    <Result
    0.08
    ond
    0.08
    æľīåĩłä¸ª
    0.08
    ä¸Ģæ¢Ŀ
    0.08
    itted
    0.08
    iad
    0.08
    Activations Density 0.033%

    No Known Activations