AeSlides: Incentivizing Aesthetic Layout in LLM-Based Slide Generation via Verifiable Rewards

• URL: http://arxiv.org/abs/2604.22840v1
• Date: Tue, 21 Apr 2026 11:59:03 GMT
• Title: AeSlides: Incentivizing Aesthetic Layout in LLM-Based Slide Generation via Verifiable Rewards
• Abstract summary: AeSlides is a reinforcement learning framework with verifiable rewards for Aesthetic layout supervision in Slide generation.<n>With only 5K training prompts on GLM-4.7-Flash, AeSlides improves aspect ratio compliance from 36% to 85%, while reducing whitespace by 44%, element collisions by 43%, and visual imbalance by 28%.
• Score: 25.706853645118453
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) have demonstrated strong potential in agentic tasks, particularly in slide generation. However, slide generation poses a fundamental challenge: the generation process is text-centric, whereas its quality is governed by visual aesthetics. This modality gap leads current models to frequently produce slides with aesthetically suboptimal layouts. Existing solutions typically rely either on heavy visual reflection, which incurs high inference cost yet yields limited gains; or on fine-tuning with large-scale datasets, which still provides weak and indirect aesthetic supervision. In contrast, the explicit use of aesthetic principles as supervision remains unexplored. In this work, we present AeSlides, a reinforcement learning framework with verifiable rewards for Aesthetic layout supervision in Slide generation. We introduce a suite of meticulously designed verifiable metrics to quantify slide layout quality, capturing key layout issues in an accurate, efficient, and low-cost manner. Leveraging these verifiable metrics, we develop a GRPO-based reinforcement learning method that directly optimizes slide generation models for aesthetically coherent layouts. With only 5K training prompts on GLM-4.7-Flash, AeSlides improves aspect ratio compliance from 36% to 85%, while reducing whitespace by 44%, element collisions by 43%, and visual imbalance by 28%. Human evaluation further shows a substantial improvement in overall quality, increasing scores from 3.31 to 3.56 (+7.6%), outperforming both model-based reward optimization and reflection-based agentic approaches, and even edging out Claude-Sonnet-4.5. These results demonstrate that such a verifiable aesthetic paradigm provides an efficient and scalable approach to aligning slide generation with human aesthetic preferences. Our repository is available at https://github.com/ympan0508/aeslides.

Related papers

• Learning to Present: Inverse Specification Rewards for Agentic Slide Generation [15.152850806890065]
  This work proposes an OpenEnv-compatible reinforcement learning environment where LLM agents learn to research topics, plan content, and generate professional HTML slide presentations through tool use.<n>We introduce a multi-component reward system combining structural validation, render quality assessment, LLM-based aesthetic scoring, content quality metrics, and an inverse specification reward that measures how faithfully generated slides convey their intended purpose.<n> Experiments on 48 diverse business briefs across six models demonstrate that our fine-tuned 7B model achieves 91.2% of Claude Opus 4.6's quality while improving 33.1% over the base model.
  arXiv  Detail & Related papers  (2026-03-17T17:45:53Z)
• SlidesGen-Bench: Evaluating Slides Generation via Computational and Quantitative Metrics [49.092634600260965]
  We introduce SlidesGen-Bench, a benchmark designed to evaluate slide generation through a lens of three core principles.<n>First, we ground our analysis in the visual domain, treating terminal outputs as renderings to remain agnostic to the underlying generation method.<n>Second, we propose a computational approach that quantitatively assesses slides across three distinct dimensions - Content, Aesthetics, and Editability.
  arXiv  Detail & Related papers  (2026-01-14T13:50:30Z)
• EasyOcc: 3D Pseudo-Label Supervision for Fully Self-Supervised Semantic Occupancy Prediction Models [4.347156881555099]
  Self-supervised models have recently achieved notable advancements, particularly in the domain of semantic occupancy prediction.<n>These models utilize sophisticated loss computation strategies to compensate for the absence of ground-truth labels.<n>We propose 3D pseudo-ground-truth labels generated by the foundation models Grounded-SAM and Metric3Dv2, and harness temporal information for label densification.
  arXiv  Detail & Related papers  (2025-09-30T11:01:32Z)
• Dynamic Classifier-Free Diffusion Guidance via Online Feedback [53.54876309092376]
  "One-size-all" approach fails to adapt to the diverse requirements of different prompts.<n>We introduce a framework for dynamic CFG scheduling.<n>We demonstrate the effectiveness of our approach on both small-scale models and the state-of-the-art Imagen 3.
  arXiv  Detail & Related papers  (2025-09-19T16:27:19Z)
• AutoPresent: Designing Structured Visuals from Scratch [99.766901203884]
  We benchmark end-to-end image generation and program generation methods with a variety of models.<n>We create AutoPresent, an 8B Llama-based model trained on 7k pairs of instructions paired with code for slide generation.
  arXiv  Detail & Related papers  (2025-01-01T18:09:32Z)
• Trimming the Fat: Efficient Compression of 3D Gaussian Splats through Pruning [17.097742540845672]
  "Trimming the fat" is a post-hoc gradient-informed iterative pruning technique to eliminate redundant information encoded in the model. Our approach achieves around 50$times$ compression while preserving performance similar to the baseline model, and is able to speed-up computation up to 600 FPS.
  arXiv  Detail & Related papers  (2024-06-26T09:57:55Z)
• CounterCurate: Enhancing Physical and Semantic Visio-Linguistic Compositional Reasoning via Counterfactual Examples [34.71588837946776]
  We propose CounterCurate, a framework to improve visio-linguistic compositional reasoning. In particular, we identify two critical under-explored problems: the neglect of the physically grounded reasoning. We first spotlight the near-chance performance of multimodal models like CLIP and LLaVA in physically grounded compositional reasoning. We then apply simple data augmentation using grounded image generation model GLIGEN to generate fine-tuning data, resulting in significant performance improvements.
  arXiv  Detail & Related papers  (2024-02-20T18:59:55Z)
• Sample Less, Learn More: Efficient Action Recognition via Frame Feature Restoration [59.6021678234829]
  We propose a novel method to restore the intermediate features for two sparsely sampled and adjacent video frames. With the integration of our method, the efficiency of three commonly used baselines has been improved by over 50%, with a mere 0.5% reduction in recognition accuracy.
  arXiv  Detail & Related papers  (2023-07-27T13:52:42Z)
• Getting More Juice Out of Your Data: Hard Pair Refinement Enhances Visual-Language Models Without Extra Data [122.282521548393]
  Contrastive Language-Image Pre-training (CLIP) has become the standard for cross-modal image-text representation learning.<n>We introduce HELIP, a cost-effective strategy that improves CLIP models by exploiting challenging text-image pairs within existing datasets in continuous training.
  arXiv  Detail & Related papers  (2023-05-09T07:00:17Z)
• Learning Customized Visual Models with Retrieval-Augmented Knowledge [104.05456849611895]
  We propose REACT, a framework to acquire the relevant web knowledge to build customized visual models for target domains. We retrieve the most relevant image-text pairs from the web-scale database as external knowledge, and propose to customize the model by only training new modualized blocks while freezing all the original weights. The effectiveness of REACT is demonstrated via extensive experiments on classification, retrieval, detection and segmentation tasks, including zero, few, and full-shot settings.
  arXiv  Detail & Related papers  (2023-01-17T18:59:06Z)
• Toward Fast, Flexible, and Robust Low-Light Image Enhancement [87.27326390675155]
  We develop a new Self-Calibrated Illumination (SCI) learning framework for fast, flexible, and robust brightening images in real-world low-light scenarios. Considering the computational burden of the cascaded pattern, we construct the self-calibrated module which realizes the convergence between results of each stage. We make comprehensive explorations to SCI's inherent properties including operation-insensitive adaptability and model-irrelevant generality.
  arXiv  Detail & Related papers  (2022-04-21T14:40:32Z)

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.