© 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. 35495
    Prev
    Next
    INDEX
    Explanations

    advantage taken

    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
    ä¸į好æĦıæĢĿ
    -0.10
    æĢİä¹Īçľĭ
    -0.10
    ä»Ģä¹Īåľ°æĸ¹
    -0.10
    åĩłåIJį
    -0.09
     premise
    -0.09
    彪
    -0.09
    åĻľ
    -0.09
    éĤĪ
    -0.08
     Prem
    -0.08
    ä¸įèµ·
    -0.08
    POSITIVE LOGITS
     advantage
    0.15
    ä¼ĺåĬ¿
    0.13
    çļĦä¼ĺåĬ¿
    0.11
     Advantage
    0.11
    Õ¨
    0.09
     LTC
    0.09
     benefit
    0.09
     advantages
    0.09
     taken
    0.08
    èĭı
    0.08
    Activations Density 0.021%

    No Known Activations