© 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. 26942
    Prev
    Next
    INDEX
    Explanations

    attention to detail

    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¸¤å¼ł
    -0.09
    miss
    -0.09
    çĽĴ
    -0.08
    家çĶµ
    -0.08
    sla
    -0.08
    ä¸Ģ声
    -0.08
    宵
    -0.08
    çļĦåŃ¦ä¹ł
    -0.08
    èĤ©è´Ł
    -0.08
    åĮ¿åIJį
    -0.08
    POSITIVE LOGITS
     attention
    0.10
    ë°Ľ
    0.10
    çģ¾åĮº
    0.10
    迤
    0.09
    Attention
    0.09
    çĤ·
    0.09
     Attention
    0.09
    çĤ¹ç¼Ģ
    0.09
    éĢł
    0.09
     detail
    0.08
    Activations Density 0.011%

    No Known Activations