© 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. Qwen3-4B
    5. 22-TRANSCODER-HP
    7. 131171
    Prev
    Next
    INDEX
    Explanations

    contrast

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    uss
    -0.28
    aza
    -0.28
    ances
    -0.28
    æĪijä¸įæĥ³
    -0.27
    ä¸įåIJĮæĦı
    -0.27
    è¿ĺä¸įå¤Ł
    -0.27
    ä¸į让
    -0.26
     unresolved
    -0.26
     tidak
    -0.26
    anced
    -0.25
    POSITIVE LOGITS
    åıĪ被
    0.26
     Croat
    0.25
     interven
    0.25
     specifically
    0.24
    éģĹçķĻ
    0.24
    æıIJ
    0.24
    Stroke
    0.24
    å¹³åĿĩæ°´å¹³
    0.24
    åIJij社ä¼ļ
    0.24
    æĮī
    0.24
    Activations Density 0.003%

    No Known Activations