© 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. 53591
    Prev
    Next
    INDEX
    Explanations

    Evans and Evan

    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
    urg
    -0.10
    smith
    -0.09
    ×ķת
    -0.09
    éĶĿ
    -0.09
    bie
    -0.09
    VL
    -0.09
    诲
    -0.08
    _sta
    -0.08
    nat
    -0.08
    erva
    -0.08
    POSITIVE LOGITS
    ãģĮãĤī
    0.11
    èĩªå¦Ĥ
    0.10
    åİ¢
    0.10
    ãģ¾ãģĻãģĮ
    0.10
    é¢Ŀ
    0.09
    åѤåĦ¿
    0.09
    é¬ĵ
    0.09
    ilate
    0.09
    elli
    0.09
    ese
    0.09
    Activations Density 0.004%

    No Known Activations