ACL Findings2026
Design First, Code Later: Aesthetically Pleasing Template-Free Slides Generation
Zhiyao Cui, Chenxu Wang, Shuyue Hu, Yiqun Zhang, Wenqi Shao, Qiaosheng Zhang, Zhen Wang
Abstract
Producing presentation slides automatically entails coordinating narrative structure with page-level graphic design under strict spatial constraints. For such structured multimodal tasks, a well-organized design process is essential to ensure the final quality of slides. Existing approaches rely on fixed templates or directly emit executable code, thereby both limiting the creative layout-design capabilities of LLMs and bypassing the essential slide-page design step. To address these limitations, this paper (1) proposes a hierarchical slides generation workflow, DeepSlides, that systematically organizes slide design tasks without any predefined template or style, decoupling slide-page design from implementation; (2) introduces SlideDesign, a dataset tailored specifically for slides generation tasks; and (3) presents a multi-agent reinforcement learning training paradigm and trains a couple of models, SlideQwens, for slide design and implementation. Experimental results demonstrate that our proposed framework outperforms baseline methods on evaluated metrics and achieves superior performance in human preference evaluations. The dataset and code are available at https://github.com/sxswz213/DeepSlides.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
- Presenting a Paper is an Art: Self-Improvement Aesthetic Agents for Academic Presentations cites PPTAgent: Generating and Evaluating Presentations Beyond Text-to-Slides
- PPTAgent: Generating and Evaluating Presentations Beyond Text-to-Slides cites DOC2PPT: Automatic Presentation Slides Generation from Scientific Documents
- PPTAgent: Generating and Evaluating Presentations Beyond Text-to-Slides cites PPTC Benchmark: Evaluating Large Language Models for PowerPoint Task Completion
Builds on10
- DOC2PPT: Automatic Presentation Slides Generation from Scientific DocumentsTsu-Jui Fu, William Yang Wang, Daniel McDuff, Yale SongAAAI 2022 · 83 citations
- ReMA: Learning to Meta-Think for LLMs with Multi-agent Reinforcement LearningZiyu Wan, Yunxiang Li, Xiaoyu Wen, Yan Song et al.NeurIPS 2025 · 76 citations
- Presenting a Paper is an Art: Self-Improvement Aesthetic Agents for Academic PresentationsChengzhi Liu, Yuzhe YANG, Kaiwen Zhou, Zhen Zhang et al.ICLR 2026 · 14 citations
- PPTC Benchmark: Evaluating Large Language Models for PowerPoint Task CompletionYiduo Guo, Zekai Zhang, Yaobo Liang, Dongyan Zhao et al.ACL Findings 2024 · 6 citations
- PPTAgent: Generating and Evaluating Presentations Beyond Text-to-SlidesHao Zheng, Xinyan Guan, Hao Kong, Wenkai Zhang et al.EMNLP 2025 · 3 citations
Related papers
- Talk to Your Slides: High-Efficiency Slide Editing via Language-Driven Structured Data ManipulationKyudan Jung, Hojun Cho, Jooyeol Yun, Soyoung Yang et al.ACL Findings 2026
- SlideTailor: Personalized Presentation Slide Generation for Scientific PapersWenzheng Zeng, Mingyu Ouyang, Langyuan Cui, Hwee Tou NgAAAI 2026 · 3 citations
- PreGenie: An Agentic Framework for High-quality Visual Presentation GenerationXiaojie Xu, Xinli Xu, Sirui Chen, Haoyu Chen et al.EMNLP Findings 2025 · 2 citations
- DeepPresenter: Environment-Grounded Reflection for Agentic Presentation GenerationHao Zheng, Guozhao Mo, Xinru Yan, Qianhao Yuan et al.ACL Findings 2026
- Automatic Slide Updating with User-Defined Dynamic Templates and Natural Language InstructionsKun Zhou, Jiakai He, Wenmian Yang, Zhensheng Wang et al.ACL Findings 2026
