© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 36539
    Prev
    Next
    INDEX
    Explanations

    isn't

    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
    able
    -0.10
    ãģĹãģ¾ãģĦ
    -0.10
    ä¸įåĩºæĿ¥
    -0.10
    æĪijä¸Ģ缴
    -0.10
    ãģ¾ãģĹãģ¦
    -0.09
    åIJİæŀľ
    -0.09
    lew
    -0.09
    oen
    -0.09
     trầm
    -0.09
    彬
    -0.09
    POSITIVE LOGITS
    egration
    0.10
    tings
    0.10
    cks
    0.09
    褶
    0.09
     even
    0.08
    ÑĤен
    0.08
    IZER
    0.08
    ough
    0.08
    ppe
    0.08
     quite
    0.08
    Activations Density 0.076%

    No Known Activations