© 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. 10001
    Prev
    Next
    INDEX
    Explanations

    performance improvement

    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
    æijŀ
    -0.09
    好ä¹ħ
    -0.08
    ulture
    -0.08
    çļĦ社ä¼ļ
    -0.08
    å®ģæĦ¿
    -0.08
    çĬ°
    -0.08
    fits
    -0.08
    anim
    -0.08
    MSN
    -0.08
    æľīæĦıæĢĿ
    -0.08
    POSITIVE LOGITS
    ä½ĵéªĮ
    0.11
    æĢ§èĥ½
    0.11
     performance
    0.10
    é«Ķé©Ĺ
    0.09
    妥åįı
    0.09
    performance
    0.09
    寡
    0.09
     Performance
    0.09
    æķĪæŀľ
    0.09
    Performance
    0.08
    Activations Density 0.121%

    No Known Activations