© 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. 3814
    Prev
    Next
    INDEX
    Explanations

    Curcumin, Artemisinin, caffeine, Forskolin

    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
    alls
    -0.09
    rades
    -0.09
    bilt
    -0.08
    íĥķ
    -0.08
     Gad
    -0.08
    iliar
    -0.08
    çļĦä¸Ģéĥ¨åĪĨ
    -0.08
    UGE
    -0.08
     Crash
    -0.08
    åĵŃ
    -0.08
    POSITIVE LOGITS
    娣
    0.09
    åįĹå±±
    0.08
    -based
    0.08
    ä¸Ģ家
    0.08
     íķ¨
    0.08
    æijĦåħ¥
    0.08
    ç²ī
    0.08
    è¯ī
    0.08
    åĴĮå°ļ
    0.08
    ç¼µ
    0.08
    Activations Density 0.990%

    No Known Activations