© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. MLP Neurons - Layer 21
    7. 9090
    Prev
    Next
    Layer 20Layer 22
    INDEX
    Compare
    Explanations
    No Explanations Found
    Auto-Interp Tools
    Top Features by Cosine Similarity
    Embeds
    IFrame
    Link
    Negative Logits
    eworld
    -0.38
    แท
    -0.36
     искус
    -0.33
    扁
    -0.33
    enville
    -0.33
    少なく
    -0.33
    当时的
    -0.32
    원
    -0.31
    ournal
    -0.30
    刁
    -0.30
    lá
    -0.30
    新世纪
    -0.30
    示意
    -0.30
    olatile
    -0.30
    ollen
    -0.30
    plat
    -0.30
    onen
    -0.30
    想办法
    -0.29
    培养学生
    -0.29
     presidential
    -0.29
    Positive Logits
    俱
    0.32
    批量
    0.31
    撂
    0.29
    衮
    0.29
    备
    0.28
    erra
    0.28
    崽
    0.27
    idy
    0.27
    size
    0.27
    蕞
    0.27
    耕耘
    0.27
    降温
    0.26
    罅
    0.26
    淝
    0.26
    serious
    0.26
    基建
    0.25
     Budd
    0.25
    orge
    0.25
     Yin
    0.25
    磅
    0.25
    Activations Density 52.489%
    Loading dashboard...