© 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. 48266
    Prev
    Next
    INDEX
    Explanations

    composition

    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
    çļĦåĬŀæ³ķ
    -0.09
    à¸Īำ
    -0.09
    resses
    -0.09
    pig
    -0.09
    -worthy
    -0.09
    orp
    -0.09
    éĽĨä½ĵ
    -0.08
    åľºéĿ¢
    -0.08
    ä»Ķ
    -0.08
    æĬł
    -0.08
    POSITIVE LOGITS
    åıĺäºĨ
    0.11
    enance
    0.09
    å¾Īæĥ³
    0.09
     Composition
    0.09
    å®ł
    0.09
    ä¸įèī¯
    0.08
    åıijçĶŁåıĺåĮĸ
    0.08
    ÑĢÑĥÑĤ
    0.08
    æĥ³å¿ħ
    0.08
    æĥ³è¿ĩ
    0.08
    Activations Density 0.177%

    No Known Activations