© 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. 48212
    Prev
    Next
    INDEX
    Explanations

    gets worse before it gets better

    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
    akh
    -0.11
    é¢Ĥ
    -0.09
    çķıæĥ§
    -0.08
    burgh
    -0.08
     exaggerated
    -0.08
     bunk
    -0.08
    zi
    -0.08
    çĶ³
    -0.08
     nto
    -0.08
    пÑĥÑģ
    -0.08
    POSITIVE LOGITS
    è¡Ģ
    0.09
     further
    0.09
    è¶ĬæĿ¥è¶Ĭé«ĺ
    0.09
     далÑĮÑĪе
    0.09
    æĦĪåĬł
    0.09
    æĿĤ
    0.09
    è¿Ľä¸ĢæŃ¥
    0.09
    ÙĨÙĬÙĨ
    0.09
     Auss
    0.09
    è¶Ĭé«ĺ
    0.09
    Activations Density 0.064%

    No Known Activations