Transformer encoder–decoder architecture

提示词

Landscape 16:9 academic concept figure of the Transformer encoder-decoder architecture, NeurIPS camera-ready style. Two vertical column stacks side-by-side with a dashed divider.

LEFT column header: "ENCODER (×N)". Blocks bottom-to-top: "Input tokens" → "Input Embedding" → "+ Positional Encoding" → dashed "Encoder layer" containing "Multi-Head Self-Attention", "Add & Norm", "Feed-Forward", "Add & Norm", with thin curved residual arrows around each sublayer.

RIGHT column header: "DECODER (×N)". Blocks bottom-to-top: "Output tokens (shifted right)" → "Output Embedding" → "+ Positional Encoding" → dashed "Decoder layer" containing "Masked Multi-Head Self-Attention", "Add & Norm", "Multi-Head Cross-Attention" (horizontal arrow from encoder top labeled "keys, values"), "Add & Norm", "Feed-Forward", "Add & Norm". Above decoder: "Linear", "Softmax", "Output probabilities".

Title: "Transformer: encoder–decoder with multi-head attention". Subtitle: "Vaswani et al., 2017".

中文译文

16:9 横版学术概念图，展示 Transformer 编码器-解码器架构，NeurIPS 定稿（camera-ready）风格。两个竖直列堆叠并排摆放，中间用虚线分隔。

左列标题：“编码器（×N）”。方块自下而上依次为：“输入 token”→“输入嵌入”→“＋位置编码”→虚线框“编码器层”，其中包含“多头自注意力”、“Add & Norm”、“前馈网络”、“Add & Norm”，每个子层周围有细的弯曲残差箭头。

右列标题：“解码器（×N）”。方块自下而上依次为：“输出 token（右移）”→“输出嵌入”→“＋位置编码”→虚线框“解码器层”，其中包含“掩码多头自注意力”、“Add & Norm”、“多头交叉注意力”（来自编码器顶部的水平箭头，标注为“键、值”）、“Add & Norm”、“前馈网络”、“Add & Norm”。解码器上方为：“线性层”、“Softmax”、“输出概率”。

标题：“Transformer：带多头注意力的编码器–解码器架构”。副标题：“Vaswani et al., 2017”。

溯源

原作者

wuyoscar

收录来源

wuyoscar/GPT-Image2-Skill · 80 · MIT