Personalization as Inverse Planning: Learning Latent Design Intents for Agentic Slide Generation via Structural Denoising

• URL: http://arxiv.org/abs/2607.00407v1
• Date: Wed, 01 Jul 2026 04:05:47 GMT
• Title: Personalization as Inverse Planning: Learning Latent Design Intents for Agentic Slide Generation via Structural Denoising
• Abstract summary: We propose a principled framework to solve Page-level Slide Personalization (PSP) approximately.<n>By intentionally corrupting the visual structures of clean slides, SPIRE creates a verifiable task to denoise the corruption.<n>We present a proof that structural denoising is a consistent surrogate for PSP, and that the multi-agent formulation strictly reduces policy gradient variance in RL.
• Score: 31.799110459149535
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Slide design requires personalizing both deck themes and page layouts. Yet, current AI agent-based methods struggle with fine-grained, page-level design. Solely relying on prespecified templates or user verbose instructions, they fail to capture latent design intents, leaving Page-level Slide Personalization (PSP) unresolved. To close this gap, this work formulates PSP as an inverse planning problem. We propose to learn a design intent without assuming any knowledge of the specific executing tools (e.g., PowerPoint, Beamer) being used. However, relinquishing control over these tools makes the problem intractable to optimize end-to-end. To overcome this, we propose SPIRE, a principled framework to solve PSP approximately. By intentionally corrupting the visual structures of clean slides, SPIRE creates a verifiable task to denoise the corruption, whereby two agents learn to collaboratively refine executable designs via reinforcement learning (RL). We present a proof that structural denoising is a consistent surrogate for PSP, and that the multi-agent formulation strictly reduces policy gradient variance in RL. Extensive experiments demonstrate the superiority of SPIRE.

Related papers

• Design First, Code Later: Aesthetically Pleasing Template-Free Slides Generation [37.32710542829869]
  This paper proposes a hierarchical slides generation workflow, DeepSlides, that organizes slide design tasks without any predefined template or style.<n> Experimental results demonstrate that our proposed framework outperforms baseline methods on evaluated metrics and achieves superior performance in human preference evaluations.
  arXiv  Detail & Related papers  (2026-04-08T06:50:37Z)
• DesignAsCode: Bridging Structural Editability and Visual Fidelity in Graphic Design Generation [21.789797121277896]
  DesignAsCode is a novel framework that reimagines graphic design as a programmatic task using HTML/CSS.<n>DesignAsCode significantly outperforms state-of-the-art rendering baselines in both structural validity and aesthetic quality.
  arXiv  Detail & Related papers  (2026-02-06T05:10:19Z)
• SEGA: A Stepwise Evolution Paradigm for Content-Aware Layout Generation with Design Prior [30.770509440256973]
  We introduce SEGA, a novel Stepwise Evolution Paradigm for Content-Aware Layout Generation.<n>Inspired by the systematic mode of human thinking, SEGA employs a hierarchical reasoning framework with a coarse-to-fine strategy.<n>We present GenPoster-100K that is a new large-scale poster dataset with rich meta-information annotation.
  arXiv  Detail & Related papers  (2025-10-17T15:36:26Z)
• Purifying Task Vectors in Knowledge-Aware Subspace for Model Merging [83.5273168208788]
  Model merging aims to integrate task-specific abilities from individually fine-tuned models into a single model without extra training.<n>The merged model often suffers from notable performance degradation due to the conflicts caused by task-irrelevant redundancy in task vectors.<n>We propose Purifying TAsk Vectors (PAVE) in knowledge-aware subspace to overcome these challenges.
  arXiv  Detail & Related papers  (2025-10-16T14:02:57Z)
• Personalized Vision via Visual In-Context Learning [62.85784251383279]
  We present a visual in-context learning framework for personalized vision.<n>PICO infers the underlying transformation and applies it to new inputs without retraining.<n>We also propose an attention-guided seed scorer that improves reliability via efficient inference scaling.
  arXiv  Detail & Related papers  (2025-09-29T17:58:45Z)
• Improving Large Language Models Function Calling and Interpretability via Guided-Structured Templates [56.73907811047611]
  Large language models (LLMs) have demonstrated strong reasoning and tool-use capabilities.<n>LLMs often fail in real-world tool-interactions due to incorrect parameterization, poor tool selection, or misinterpretation of user intent.<n>We introduce a curriculum-inspired framework that leverages structured reasoning templates to guide LLMs through more deliberate step-by-step instructions for generating function callings.
  arXiv  Detail & Related papers  (2025-09-22T17:55:14Z)
• DesignLab: Designing Slides Through Iterative Detection and Correction [32.29179295296639]
  We propose DesignLab, which separates the design process into two roles, the design reviewer and the design contributor.<n>This decomposition enables an iterative loop where the reviewer continuously detects issues and the contributor corrects them.<n>Our experiments show that DesignLab outperforms existing design-generation methods, including a commercial tool.
  arXiv  Detail & Related papers  (2025-07-23T04:49:48Z)
• Rethinking Layered Graphic Design Generation with a Top-Down Approach [76.33538798060326]
  Graphic design is crucial for conveying ideas and messages. Designers usually organize their work into objects, backgrounds, and vectorized text layers to simplify editing.<n>With the rise of GenAI methods, an endless supply of high-quality graphic designs in pixel format has become more accessible.<n>Despite this, non-layered designs still inspire human designers, influencing their choices in layouts and text styles, ultimately guiding the creation of layered designs.<n>Motivated by this observation, we propose Accordion, a graphic design generation framework taking the first attempt to convert AI-generated designs into editable layered designs.
  arXiv  Detail & Related papers  (2025-07-08T02:26:08Z)
• Compositional Generative Inverse Design [69.22782875567547]
  Inverse design, where we seek to design input variables in order to optimize an underlying objective function, is an important problem. We show that by instead optimizing over the learned energy function captured by the diffusion model, we can avoid such adversarial examples. In an N-body interaction task and a challenging 2D multi-airfoil design task, we demonstrate that by composing the learned diffusion model at test time, our method allows us to design initial states and boundary shapes.
  arXiv  Detail & Related papers  (2024-01-24T01:33:39Z)
• Improving Long-Horizon Imitation Through Instruction Prediction [93.47416552953075]
  In this work, we explore the use of an often unused source of auxiliary supervision: language. Inspired by recent advances in transformer-based models, we train agents with an instruction prediction loss that encourages learning temporally extended representations that operate at a high level of abstraction. In further analysis we find that instruction modeling is most important for tasks that require complex reasoning, while understandably offering smaller gains in environments that require simple plans.
  arXiv  Detail & Related papers  (2023-06-21T20:47:23Z)
• Goal-Directed Design Agents: Integrating Visual Imitation with One-Step Lookahead Optimization for Generative Design [0.0]
  This note builds on DLAgents to develop goal-directed agents capable of enhancing learned strategies for sequentially generating designs. Goal-directed DLAgents can employ human strategies learned from data along with optimizing an objective function. This illustrates a design agent framework that can efficiently use feedback to not only enhance learned design strategies but also adapt to unseen design problems.
  arXiv  Detail & Related papers  (2021-10-07T07:13:20Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.