© 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. 55533
    Prev
    Next
    INDEX
    Explanations

    full presentation slides or PDF

    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
    ãģ¨ãģª
    -0.08
    æīĢæıIJä¾ĽçļĦ
    -0.08
    å¼Ħ
    -0.08
    ä¾
    -0.08
    ela
    -0.08
    æĬ¢éĻ©
    -0.08
    OMIC
    -0.08
     contr
    -0.08
     Vick
    -0.08
    对åħ¶
    -0.08
    POSITIVE LOGITS
    å®Įæķ´çļĦ
    0.12
    å®Įæķ´
    0.12
    åħ¨æĸĩ
    0.11
     full
    0.10
    æĮ¹
    0.10
    éĽī
    0.09
    full
    0.09
    åİŁæĸĩ
    0.09
    é£ŀè·ĥ
    0.09
    ErrorException
    0.09
    Activations Density 0.172%

    No Known Activations