© 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. 63671
    Prev
    Next
    INDEX
    Explanations

    all I need to do is

    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
    çļĦæĺ¯
    -0.12
     вÑģего
    -0.11
    åħ³éĶ®æĺ¯
    -0.11
    ãģĵãģĿ
    -0.10
    åıªéľĢ
    -0.10
    åģļçļĦäºĭ
    -0.10
    å¾Īéĩįè¦ģçļĦ
    -0.10
    ä¸įåıªæĺ¯
    -0.10
    çļĦåħ³éĶ®
    -0.10
    ãģ®ãģ¯
    -0.09
    POSITIVE LOGITS
    (er
    0.10
    matic
    0.10
    åĪ°å¤Ħ
    0.09
    omers
    0.08
    æĬĢæľ¯åĴĮ
    0.08
     Kane
    0.08
    KN
    0.08
    èĿī
    0.08
    æ¦Ħ
    0.08
    Ùħا
    0.08
    Activations Density 0.018%

    No Known Activations