2048 × 1152原文件 · 2.8 MiB
视觉设计 · 论文插图 · 外部收藏
语言模型角色图谱
提示词
Create a premium conceptual figure for an EMNLP / ACL paper, landscape 16:9, high-resolution, polished editorial-academic style. Theme: "LLM Persona Atlas". This should not look like a generic pipeline diagram. It should look like a beautifully designed Figure 1 from a top NLP / agent paper: minimal, refined, memorable, with a strong central visual metaphor.
Use a warm off-white paper background, subtle grain, large clean margins, crisp vector-like linework, delicate shadows, and fine gradients used sparingly. Use an understated, high-end color palette: ink black, warm gray, muted cobalt, dusty teal, soft sage, pale amber, muted coral, slate blue. No saturated rainbow colors, no cartoon style, no photorealism, no generic stock illustration.
Composition: left "Utterance Stream" with small translucent speech fragments flowing in as curved data ribbons; center "Persona Lens" as a glass-like hexagonal prism / agent lens that refracts utterance ribbons into six colored persona strands; right "Six Persona Glyphs" as a coherent 2x3 gallery of abstract symbolic avatars labeled "Concise", "Explainer", "Cautious", "Supportive", "Creative", and "Analyst".
Keep typography sparse, crisp, and clean. Add a small title "LLM Persona Atlas" and subtitle "from utterance style to model profile". Avoid dense method labels, big boxes, fake equations, fake citations, garbled text, photoreal humans, childish cartoon avatars, heavy shadows, and purple gradient backgrounds.
原始生成信息
- 平台
- 未记录
- 模型显示名
- 未记录模型
- 模型 ID
- 未记录
- 模型证据
- unknown
- 生成时间
- 未记录
来源与使用说明
- 作者
- Wuyoscar(整理)
- 贡献者
- wuyoscar/GPT-Image2-Skill
- 来源
- 查看原始来源
- Prompt 许可
- unspecified
- 媒体许可
- unspecified
收录自 GPT-Image2-Skill(上游提交 05cb1130bba29e0fc028220376280a2e934a8041,案例 No. 94)。上游仓库声明 MIT License,并要求保留图库作者和来源标注;部分案例另有第三方作者/来源,原始素材的单独授权未核实。仓库许可全文见本站项目 docs/third-party/GPT-Image2-Skill-LICENSE.txt。源图库:https://github.com/wuyoscar/GPT-Image2-Skill/blob/05cb1130bba29e0fc028220376280a2e934a8041/skills/gpt-image/references/gallery-research-paper-figures.md