© 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. 107474
    Prev
    Next
    INDEX
    Explanations

    Excess or quantity

    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
     universe
    -0.26
    icks
    -0.26
    å®ľ
    -0.26
    storage
    -0.26
    åĮĹæĸĹ
    -0.24
    ÑĤÑĥÑĢ
    -0.24
    esh
    -0.24
     Universe
    -0.24
    åıĸèĥľ
    -0.24
    UDA
    -0.23
    POSITIVE LOGITS
     inexperienced
    0.28
    æ³³
    0.26
     em
    0.24
    æĥħåĨµä¸ĭ
    0.24
    lessly
    0.24
     ham
    0.24
     sailed
    0.24
    æ²Ļ
    0.24
    antly
    0.24
     sand
    0.24
    Activations Density 3.864%

    No Known Activations