© 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. 3677
    Prev
    Next
    INDEX
    Explanations

    difficult job or challenge

    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¸ĩä¸Ģ
    -0.10
    ä¸Ģçľ¼
    -0.10
    便äºİ
    -0.09
    驹
    -0.09
    æĽ´å¥½
    -0.09
    æĹ©æĹ¥
    -0.09
    ioni
    -0.09
    æ±
    -0.08
    涧
    -0.08
    ugins
    -0.08
    POSITIVE LOGITS
    éļ¾åº¦
    0.15
    åĽ°éļ¾
    0.14
    åĽ°éĽ£
    0.13
    æĺ¯ä¸Ģ项
    0.12
    æĮijæĪĺ
    0.12
    èī°å·¨
    0.11
     challenging
    0.11
     challenge
    0.11
    challenge
    0.11
     difficulty
    0.11
    Activations Density 0.129%

    No Known Activations