© 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. 108234
    Prev
    Next
    INDEX
    Explanations

    numerous

    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
    éĽĸ
    -0.30
    æĭĮ
    -0.27
    æĹ¢æĺ¯
    -0.26
    eri
    -0.25
    инÑĤеÑĢ
    -0.25
    emens
    -0.24
    ivated
    -0.24
    广大
    -0.24
    èĻ½
    -0.24
    çĭ¬çī¹çļĦ
    -0.23
    POSITIVE LOGITS
     number
    0.30
    filled
    0.29
     single
    0.28
     par
    0.28
     non
    0.28
     trail
    0.27
     reass
    0.27
     form
    0.27
     con
    0.26
     mis
    0.26
    Activations Density 11.274%

    No Known Activations