© 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. 10973
    Prev
    Next
    INDEX
    Explanations

    need to know why

    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
    æĶĢçĻ»
    -0.10
     дал
    -0.10
    æľī两个
    -0.09
    åĪļåĪļ
    -0.09
    足足
    -0.09
    èķŀ
    -0.08
    ä½łæĥ³
    -0.08
    æľĢåĪĿ
    -0.08
    å¤ļåįĬ
    -0.08
    æ¸´æľĽ
    -0.08
    POSITIVE LOGITS
    æĪĸèĢħ说
    0.10
    ãĤĦãģ¯ãĤĬ
    0.10
    åĵªäºĽ
    0.10
    ä¹ĭç±»çļĦ
    0.10
    åīįè¾Ī
    0.09
    .kind
    0.09
    ä¼¼çļĦ
    0.09
    Tot
    0.09
    ĪæĿĥ
    0.08
    仲
    0.08
    Activations Density 1.429%

    No Known Activations