© 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. Qwen3-4B
    5. 20-TRANSCODER-HP
    7. 63856
    Prev
    Next
    INDEX
    Explanations

    to

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    人为
    -0.33
    åĨįåĪ°
    -0.30
    ohen
    -0.28
    umer
    -0.28
    åĽłä¸ºä½ł
    -0.26
    idian
    -0.25
    沿çĿĢ
    -0.25
    centaje
    -0.24
     Memo
    -0.24
    för
    -0.24
    POSITIVE LOGITS
    èµİ
    0.30
    Liquid
    0.25
    coal
    0.25
    ä¸Ģ个å¤ļ
    0.24
     triples
    0.24
    åIJĮèĥŀ
    0.24
    æĽ´éĩįè¦ģçļĦ
    0.24
    å¿®
    0.24
     thr
    0.24
    /bg
    0.24
    Activations Density 0.986%

    No Known Activations