Narrative-Driven Paper-to-Slide Generation via ArcDeck
Summary
ArcDeck is a multi-agent framework that generates presentation slides from academic papers by modeling logical flow through discourse trees and iterative agent refinement, outperforming direct summarization methods. The paper introduces ArcBench, a new benchmark for evaluating paper-to-slide generation with emphasis on narrative coherence and logical structure.
View Cached Full Text
Cached at: 04/20/26, 08:29 AM
Paper page - Narrative-Driven Paper-to-Slide Generation via ArcDeck
Source: https://huggingface.co/papers/2604.11969
Abstract
ArcDeck is a multi-agent framework that enhances paper-to-slide generation by modeling logical flow through discourse trees and iterative agent refinement, outperforming direct summarization methods.
We introduce ArcDeck, a multi-agent framework (https://huggingface.co/papers?q=multi-agent%20framework) that formulates paper-to-slide generation as a structured narrative reconstruction (https://huggingface.co/papers?q=narrative%20reconstruction) task. Unlike existing methods that directly summarize raw text into slides, ArcDeck explicitly models the source paper’s logical flow. It first parses the input to construct a discourse tree (https://huggingface.co/papers?q=discourse%20tree) and establish a global commitment document (https://huggingface.co/papers?q=global%20commitment%20document), ensuring the high-level intent is preserved. These structural priors then guide an iterative multi-agent refinement (https://huggingface.co/papers?q=iterative%20multi-agent%20refinement) process, where specialized agents iteratively critique and revise the presentation outline before rendering the final visual layouts and designs. To evaluate our approach, we also introduce ArcBench, a newly curated benchmark of academic paper-slide pairs. Experimental results demonstrate that explicit discourse modeling, combined with role-specific agent coordination, significantly improves the narrative flow and logical coherence of the generated presentations.
View arXiv page (https://arxiv.org/abs/2604.11969)View PDF (https://arxiv.org/pdf/2604.11969)Project page (https://arcdeck.org/)GitHub40 (https://github.com/RehgLab/ArcDeck)Add to collection (https://huggingface.co/login?next=%2Fpapers%2F2604.11969)
Get this paper in your agent:
hf papers read 2604.11969
Don’t have the latest CLI?curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2604.11969 in a model README.md to link it from this page.
Datasets citing this paper1
ArcDeck/ArcBench Viewer• Updated3 days ago • 2 • 215 (https://huggingface.co/datasets/ArcDeck/ArcBench)
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2604.11969 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to a collection (https://huggingface.co/new-collection) to link it from this page.
Similar Articles
DeepSlide: From Artifacts to Presentation Delivery
DeepSlide is a human-in-the-loop multi-agent system for the full presentation process, from requirement elicitation and time-budgeted narrative planning to evidence-grounded slide-script generation and rehearsal support. It introduces a dual-scoreboard benchmark separating static artifact quality from dynamic delivery excellence, and achieves gains in narrative flow, pacing precision, and slide-script synergy.
X+Slides: Benchmarking Audience-Conditioned Slide Generation
X+Slides is a new benchmark for evaluating audience-conditioned slide generation from source documents, using source-grounded probes and audience-specific utility weights. Experiments on DeepPresenter, SlideTailor, and NotebookLM show that current systems recover substantial but incomplete audience-essential information.
Deckwise
Deckwise is an AI presentation agent that helps users create and edit presentation decks.
I've been starting from a blank document every time I need to build a presentation. Claude can build the whole deck from a conversation and I only figured this out recently.
Describes a workflow using Claude AI to build presentations through a conversational interview process, resulting in better-structured slides than manual drafting.
I built 10 gamified, interactive presentation decks to teach Agentic AI (Stop falling asleep reading whitepapers).
A developer built 10 gamified, interactive slide decks within the AgentSwarms platform to teach Agentic AI concepts like ReAct loops, multi-agent swarms, and production RAG, using active recall instead of passive reading.