© 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. 11547
    Prev
    Next
    INDEX
    Explanations

    cereal and breakfast foods

    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
    æģģ
    -0.10
    ä¸įæŃ¢
    -0.09
    fat
    -0.08
     Underground
    -0.08
     fat
    -0.08
    èŃ°
    -0.08
    è£ģ
    -0.08
    inning
    -0.08
    æĥ¯
    -0.08
    éĤ£åıª
    -0.08
    POSITIVE LOGITS
     cereal
    0.13
     cere
    0.11
    ç³ģ
    0.10
    FirstChild
    0.10
     breakfast
    0.10
     grav
    0.09
    èĬ½
    0.09
    é¢Ĺç²Ĵ
    0.09
    ç²¥
    0.09
    å¡Į
    0.09
    Activations Density 0.023%

    No Known Activations