© 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. 1660
    Prev
    Next
    INDEX
    Explanations

    leaking transformer; bleed; leakage

    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
    æİ¥åĬĽ
    -0.10
    olv
    -0.09
    .tem
    -0.09
    åIJīåĪ©
    -0.09
    äºī
    -0.09
    éĺħ读åħ¨æĸĩ
    -0.09
     rubble
    -0.08
     consumption
    -0.08
    æīĭ表
    -0.08
    äºĨåIJĹ
    -0.08
    POSITIVE LOGITS
     bleeding
    0.13
     bleed
    0.13
     Ble
    0.12
    åĩºè¡Ģ
    0.11
    æ¼ı
    0.11
    æ¸Ĺ
    0.10
     draining
    0.09
    æ»´
    0.09
     leaking
    0.09
     leakage
    0.09
    Activations Density 0.093%

    No Known Activations