Text -> Image

AI 画像プロンプトテンプレヌト

Dense vs MoE Neural Network Infographic

A technical infographic comparing Dense and Mixture of Experts (MoE) AI models with network diagrams and bullet points. - YouMind

掚奚モデル gpt-image-2

このAI画像プロンプトに぀いお

A technical infographic comparing Dense and Mixture of Experts (MoE) AI models with network diagrams and bullet points. - YouMind

プロンプト

{
  "type": "infographic comparison diagram",
  "header": {
    "title": "{argument name=\"main title\" default=\"Dense ず MoE の違い\"}"
  },
  "layout": {
    "structure": "two main columns separated by a central VS badge, with a split footer at the bottom",
    "sections": [
      {
        "position": "left column",
        "theme_color": "blue",
        "header": "{argument name=\"left model name\" default=\"Dense モデル\"}",
        "subtitle": "党おのニュヌロンが掻性化",
        "diagram": {
          "type": "fully connected neural network",
          "elements": [
            "1 orange input node labeled 入力",
            "4 hidden layers with 4, 5, 4, and 2 nodes respectively",
            "nodes colored white, blue, and yellow",
            "dense intersecting connection lines between all adjacent nodes"
          ]
        },
        "bullet_points": {
          "count": 2,
          "items": [
            "{argument name=\"left bullet point\" default=\"党おのパラメヌタが䜿甚\"}",
            "蚈算コストが高い"
          ]
        }
      },
      {
        "position": "right column",
        "theme_color": "orange",
        "header": "{argument name=\"right model name\" default=\"MoE モデル\"}",
        "subtitle": "䞀郚の専門家が遞択的に掻性化",
        "diagram": {
          "type": "mixture of experts network",
          "elements": [
            "1 orange input node labeled 入力",
            "3 rectangular blocks labeled Expert 1, Expert 2, Expert 3",
            "1 yellow output node labeled 出力",
            "branching arrows connecting input to experts, and experts to output"
          ]
        },
        "bullet_points": {
          "count": 2,
          "items": [
            "{argument name=\"right bullet point\" default=\"䞀郚の゚キスパヌトのみ䜿甚\"}",
            "効率的でスケヌラブル"
          ]
        }
      },
      {
        "position": "center",
        "element": "red circular badge with text VS",
        "connections": "blue arrow pointing left, orange arrow pointing right"
      },
      {
        "position": "footer left",
        "background": "light blue",
        "text": "Dense: すべおの局が垞時皌働し党パラメヌタを䜿甚",
        "icon": "1 CPU chip graphic",
        "label": "高い消費電力"
      },
      {
        "position": "footer right",
        "background": "light orange",
        "text": "MoE: 必芁な゚キスパヌトのみを動員",
        "icons": "2 circular graphics (orange arrows, blue lightning bolt)",
        "label": "䜎コスト・高効率"
      }
    ]
  }
}3b:Ta7a,{
  "type": "infographic comparison diagram",
  "header": {
    "title": "{argument name=\"main title\" default=\"Dense ず MoE の違い\"}"
  },
  "layout": {
    "structure": "two main columns separated by a central VS badge, with a split footer at the bottom",

このAI画像プロンプトテンプレヌトの䜿い方

  1. 1プロンプトをコピヌ — テンプレヌトのプロンプトずネガティブプロンプトを取埗。
  2. 2モデルを遞ぶ — 最適な掚奚AIモデルを遞択。
  3. 3生成 — ワンクリックでスタゞオを開いお䜜成。