© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 9938
    Prev
    Next
    INDEX
    Explanations

    dominant, dominating, effective

    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
     Useful
    -0.10
    åħ¨åĬĽä»¥èµ´
    -0.09
    .RequestMethod
    -0.09
    æĢ¨
    -0.09
     halftime
    -0.09
    èŃ
    -0.09
    æĥ°
    -0.08
    æļ¹
    -0.08
    éī´
    -0.08
    èĬ±
    -0.08
    POSITIVE LOGITS
    ereal
    0.09
    顺
    0.09
     اÙĦاث
    0.09
    éļıæīĭ
    0.09
    åģ¶
    0.09
    å¡«è¡¥
    0.08
    hone
    0.08
    çļĦç§ĺå¯Ĩ
    0.08
    ä½ıäºĨ
    0.08
    æį®
    0.08
    Activations Density 0.186%

    No Known Activations