© 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. 107072
    Prev
    Next
    INDEX
    Explanations

    feedback

    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
     tests
    -0.29
     sv
    -0.28
    æĪIJ年人
    -0.27
    å¼Ľ
    -0.25
    Ed
    -0.24
    icks
    -0.24
    iming
    -0.24
    otto
    -0.24
     Ed
    -0.24
    ched
    -0.24
    POSITIVE LOGITS
     cess
    0.33
    åĴĮæĬĢæľ¯
    0.27
    è°¿
    0.25
    ÑģÑĤавил
    0.24
     bergen
    0.24
    /update
    0.24
    æĽī
    0.24
     smoothed
    0.24
    kim
    0.23
    .Toggle
    0.23
    Activations Density 1.691%

    No Known Activations