© 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. 26923
    Prev
    Next
    INDEX
    Explanations

    file or directory names

    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
    æŁ¥çľĭåħ¨æĸĩ
    -0.10
     <!--[
    -0.09
    èijĨ
    -0.09
     ê·¸ë¦¬ê³ł
    -0.09
    دس
    -0.08
    çĽĬ
    -0.08
    å¥
    -0.08
    çijļ
    -0.08
     riv
    -0.08
    è±ģ
    -0.08
    POSITIVE LOGITS
    agg
    0.08
    å«©
    0.08
    (inner
    0.08
     Wilson
    0.08
    Sou
    0.08
    è´¨çĸij
    0.07
    æľ¬æĿ¥
    0.07
    lke
    0.07
    SE
    0.07
    5
    0.07
    Activations Density 0.679%

    No Known Activations