© 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. 17449
    Prev
    Next
    INDEX
    Explanations

    adverbs of time and manner

    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
    æĪĸèĢħ说
    -0.10
     aba
    -0.08
    еннÑĭй
    -0.08
    顺
    -0.08
     ep
    -0.07
    æĪĸåħ¶ä»ĸ
    -0.07
    çĿ¢
    -0.07
    æĪĸ
    -0.07
    atch
    -0.07
    stood
    -0.07
    POSITIVE LOGITS
    éĵ¶è¡Įä¸ļ
    0.11
    äºĽä»Ģä¹Ī
    0.10
    çī¹ä»·
    0.10
    raphics
    0.10
    è¿Ķ乡
    0.10
    èĢģ人家
    0.10
    å¾Īå·®
    0.10
    èĮļ
    0.10
    ä¹Łæĺ¯éĿŀ常
    0.09
    å°±è·Ł
    0.09
    Activations Density 0.263%

    No Known Activations