© 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. 130680
    Prev
    Next
    INDEX
    Explanations

    C++ code symbols/keywords

    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
    çīĪæĿĥæīĢæľī
    -0.26
    æŃ¤äºº
    -0.24
    éĻ¢
    -0.24
    aguay
    -0.23
    å¸Ĥåľºä¸Ĭ
    -0.23
    é¡¹çĽ®çļĦ
    -0.23
    为ä¸ŃåĽ½
    -0.23
     GDK
    -0.23
    äh
    -0.23
    cul
    -0.23
    POSITIVE LOGITS
     singly
    0.27
     vig
    0.25
     limp
    0.25
    itious
    0.24
    byname
    0.24
    nable
    0.24
    EMU
    0.24
    اش
    0.24
     McGill
    0.24
    .services
    0.24
    Activations Density 8.598%

    No Known Activations