© 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. 39838
    Prev
    Next
    INDEX
    Explanations

    click on image or video

    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
    广大群ä¼Ĺ
    -0.09
    boss
    -0.09
    计è¾ĥ
    -0.09
    äºĨä¸Ģ份
    -0.09
    éĢłè¡Ģ
    -0.09
    uktur
    -0.08
    mind
    -0.08
    agr
    -0.08
    /content
    -0.08
    èļĵ
    -0.08
    POSITIVE LOGITS
     Harden
    0.10
     дал
    0.10
    åį³åı¯
    0.09
    以便
    0.09
    çļĦè¯Ŀé¢ĺ
    0.09
    åħ´èĩ´
    0.09
    俸
    0.09
    èĢĮæĺ¯
    0.08
    æĹģè¾¹çļĦ
    0.08
     Kok
    0.08
    Activations Density 0.069%

    No Known Activations