Reliable Reasoning in SVG-LLMs via Multi-Task Multi-Reward Reinforcement Learning

TL;DR

This paper introduces CTRL-S, a reinforcement learning framework with a chain-of-thought mechanism and multi-reward optimization, significantly improving SVG generation quality and reasoning in vision-language models.

Contribution

The paper presents a novel multi-task, multi-reward reinforcement learning approach with a chain-of-thought mechanism for SVG generation, supported by a new high-quality SVG dataset.

Findings

CTRL-S outperforms existing methods in success rates.

Improves SVG code structure and visual fidelity.

Enhances reasoning and generalization in SVG generation.

Abstract

With the rapid advancement of vision-language models, an increasing number of studies have explored their potential for SVG generation tasks. Although existing approaches improve performance by constructing large-scale SVG datasets and introducing SVG-specific tokens, they still suffer from limited generalization, redundant paths in code outputs, and a lack of explicit reasoning. In this work, we present CTRL-S (Chain-of-Thought Reinforcement Learning for SVG), a unified framework that introduces a chain-of-thought mechanism to explicitly expose the model's reasoning process during SVG generation. To support this structured reasoning, we construct SVG-Sophia, a high-quality dataset containing 145K samples across SVG code refinement, Text-to-SVG, and Image-to-SVG tasks. By training the model to generate group-level structured SVG code, CTRL-S significantly improves structural coherence and visual fidelity. Furthermore, we adopt the GRPO algorithm and design a multi-reward optimization framework, incorporating DINO, image-text similarity, format, and code efficiency rewards. Through joint multi-reward optimization and multi-task training, our approach systematically enhances overall generation capabilities. Extensive experiments show that CTRL-S outperforms existing methods, achieving higher task success rates, superior SVG code quality, and exceptional visual fidelity.