# GPT Image Hub prompt

## 다크 모드 연구 논문 도표 라이브러리 UI
- URL: https://www.gptimagehub.com/ko/library/dark-scientific-figure-library-ui
- Generator URL: https://www.gptimagehub.com/ko/generate?promptId=oss-ym-g2-23219
- Markdown URL: https://www.gptimagehub.com/api/agent/markdown?path=%2Fko%2Flibrary%2Fdark-scientific-figure-library-ui
- Category: UI / UX 디자인
- Model: gpt-image-2
- Aspect ratio: 16:9
- Style: graphic-design
- Tags: ui 디자인, 대시보드, 다크 모드, 학술 도표, 갤러리 색인, saas 인터페이스, 인포그래픽
- Updated: 2026-10-27T05:53:13.098Z
- Created: 2026-10-27T05:53:13.098Z

학술 논문 다이어그램 프리셋을 색인하는 다크 모드 웹 대시보드를 생성합니다. 대괄호 안의 제목과 카테고리를 변경하여 커스텀할 수 있습니다.

### Full prompt
```text
A polished dark-mode web application UI screenshot displaying a gallery index page for a curated research paper figure style library, titled "[研究论文图示]". Wide landscape 16:9 browser-like composition on a dark charcoal background with subtle grid lines, crisp typography, and high-contrast white and cyan accents. At the top center, the title "[研究论文图示]" appears beside a small stacked multicolor document icon, separated by a thin horizontal divider. In the upper-right corner is a compact rounded button labeled "↑ 图库索引". Below the divider on the left sits the bold section heading "[研究论文图示网格]". The main area features a 4-column grid of gallery cards with dark gutters and thin gray borders, showing 8 complete cards across the first two rows and 4 partially visible cards clipped at the bottom. Each card displays a clean, white-background scientific figure thumbnail, a bold title underneath, and three small dark tag pills reading "landscape", "high", and "Curated". Top cards include: Card 1 titled "患者队列与多模态生物标志物流称流程" with patient cohort diagrams; Card 2 titled "单细胞免疫图谱" with UMAP clusters and trajectory plots; Card 3 titled "多模态医疗 AI 方法图" with multimodal clinical AI pipeline; Card 4 titled "治疗响应统计图" with forest plots and statistics; Card 5 titled "Transformer 编码器-解码器架构" with stacked neural blocks; Card 6 titled "多智能体 LLM 系统架构" with central LLM and tool modules; Card 7 titled "去噪扩散正/逆向链" with diffusion process panels; Card 8 titled "经验缩放规律图" with scaling law curves. The bottom clipped cards begin with English titles: "Benchmark comparison across 10 frontier LLMs", "Ablation of core reasoning components across 5 benchmarks", "LLM pretraining data mixture and downstream splits", and "Representative multi-head attention patterns in a 16-layer Transformer". Clean SaaS aesthetic, no browser address bars, no photo clutter, no watermarks.
```