TILT: Improving Compositional Generation in Diffusion Models with a Model-Intrinsic Reward

• URL: http://arxiv.org/abs/2607.21606v1
• Date: Sat, 16 May 2026 16:56:52 GMT
• Title: TILT: Improving Compositional Generation in Diffusion Models with a Model-Intrinsic Reward
• Abstract summary: We present TILT, a training-free framework for compositional text-to-image generation via test-time reward alignment.<n>We interpret compositional failures as overlap modes between joint and single-concept distributions, and define a reward that favors samples where all concepts are jointly present.<n>Experiments on prompts from T2ICompBench show that our method improves compositional alignment while preserving image quality compared to previous baselines.
• Score: 7.47572316039482
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advances in powerful text-to-image generation models have made it increasingly important to develop test-time methods that modify the sampling trajectory to produce images more faithful to complex compositional prompts. We present TILT, a training-free framework for compositional text-to-image generation via test-time reward alignment. We interpret compositional failures as overlap modes between joint and single-concept distributions, and define a reward that favors samples where all concepts are jointly present. This reward is intrinsic to the base model and does not require any external supervision or reward models. This yields a KL-constrained objective with a closed-form tilted target distribution and principled guiding steps for diffusion sampling. The interaction of concept distributions together with the above reward naturally leads to two different guidance strategies while a hybrid approach that balances their respective benefits produces stronger performance. Experiments on prompts from T2ICompBench show that our method improves compositional alignment while preserving image quality compared to previous baselines.

Related papers

• Coevolving Representations in Joint Image-Feature Diffusion [13.292849179066993]
  We argue that the representation space guiding diffusion should itself adapt to the generative task.<n>We propose Coevolving Diffusion Representation (CoReDi), a framework in which the semantic representation space evolves during training.<n>Experiments show that CoReDi achieves faster convergence and higher sample quality compared to joint diffusion models operating in fixed representation spaces.
  arXiv  Detail & Related papers  (2026-04-19T15:29:15Z)
• Reward Score Matching: Unifying Reward-based Fine-tuning for Flow and Diffusion Models [54.597200388738656]
  Reward-based fine-tuning aims to steer a pretrained diffusion or flow-based generative model toward higher-reward samples.<n>We show that many can be written under a common framework, which we call reward score matching (RSM)
  arXiv  Detail & Related papers  (2026-04-19T12:47:52Z)
• UniGRPO: Unified Policy Optimization for Reasoning-Driven Visual Generation [51.41441081823758]
  Unified models capable of interleaved generation have emerged as a promising paradigm.<n>We propose a unified reinforcement learning framework tailored for interleaved generation.<n>Our experiments demonstrate that this unified training recipe significantly enhances image generation quality through reasoning.
  arXiv  Detail & Related papers  (2026-03-24T17:59:17Z)
• Alignment of Diffusion Model and Flow Matching for Text-to-Image Generation [39.484148941369234]
  Diffusion models and flow matching have demonstrated remarkable success in text-to-image generation.<n>We propose a novel alignment framework by leveraging the underlying nature of the alignment problem.<n>We achieve comparable performance to finetuning-based models with one-step generation with at least a 60% reduction in computational cost.
  arXiv  Detail & Related papers  (2026-01-31T00:06:55Z)
• Fast, faithful and photorealistic diffusion-based image super-resolution with enhanced Flow Map models [8.773008452897791]
  FlowMapSR is a novel diffusion-based framework for image super-resolution explicitly designed for efficient inference.<n>We show that FlowMapSR achieves a better balance between reconstruction faithfulness and photorealism than recent state-of-the-art methods for both x4 and x8 upscaling.
  arXiv  Detail & Related papers  (2026-01-23T11:25:04Z)
• HiCoGen: Hierarchical Compositional Text-to-Image Generation in Diffusion Models via Reinforcement Learning [66.99487505369254]
  HiCoGen is built upon a novel Chain of Synthesis paradigm.<n>It decomposes complex prompts into minimal semantic units.<n>It then synthesizes these units iteratively, where the image generated in each step provides crucial visual context for the next.<n>Experiments show our approach significantly outperforms existing methods in both concept coverage and compositional accuracy.
  arXiv  Detail & Related papers  (2025-11-25T06:24:25Z)
• Composition and Alignment of Diffusion Models using Constrained Learning [79.36736636241564]
  Diffusion models have become prevalent in generative modeling due to their ability to sample from complex distributions.<n>Two commonly used methods are: (i) alignment, which involves fine-tuning a diffusion model to align it with a reward; and (ii) composition, which combines several pre-trained diffusion models, each emphasizing a desirable attribute in the generated outputs.<n>We propose a constrained optimization framework that unifies alignment and composition of diffusion models by enforcing that the aligned model satisfies reward constraints and/or remains close to (potentially multiple) pre-trained models.
  arXiv  Detail & Related papers  (2025-08-26T15:06:30Z)
• IterComp: Iterative Composition-Aware Feedback Learning from Model Gallery for Text-to-Image Generation [70.8833857249951]
  IterComp is a novel framework that aggregates composition-aware model preferences from multiple models.<n>We propose an iterative feedback learning method to enhance compositionality in a closed-loop manner.<n>IterComp opens new research avenues in reward feedback learning for diffusion models and compositional generation.
  arXiv  Detail & Related papers  (2024-10-09T17:59:13Z)
• JoReS-Diff: Joint Retinex and Semantic Priors in Diffusion Model for Low-light Image Enhancement [69.6035373784027]
  Low-light image enhancement (LLIE) has achieved promising performance by employing conditional diffusion models. Previous methods may neglect the importance of a sufficient formulation of task-specific condition strategy. We propose JoReS-Diff, a novel approach that incorporates Retinex- and semantic-based priors as the additional pre-processing condition.
  arXiv  Detail & Related papers  (2023-12-20T08:05:57Z)

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.