© 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. 54225
    Prev
    Next
    INDEX
    Explanations

    decaying, tail, -aying

    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
    çĦ¶æĺ¯
    -0.10
    ogh
    -0.09
    åĮĸçļĦ
    -0.09
    ä¸įäºĨ
    -0.09
    anning
    -0.09
    iating
    -0.09
    èĢģ人家
    -0.09
    绣
    -0.09
    å°±è¿ŀ
    -0.09
    å±¹
    -0.09
    POSITIVE LOGITS
    è¿ĽæŃ¥
    0.09
    ä¿ļ
    0.08
    ards
    0.08
     lash
    0.08
    åIJĮå¿Ĺ们
    0.08
     Lei
    0.08
    æľªæĪIJ
    0.08
    istr
    0.07
    orts
    0.07
     gm
    0.07
    Activations Density 0.015%

    No Known Activations