AI TrackerFrontier desk · AEST
Back to news
normalaitech

AesCode: Aesthetic Code Generation with Decoupled Cross-Modal Rewards

arXiv cs.AI · 14 Oct 2026, 03:59 AEST

AesCode introduces a novel approach to generating aesthetic code for visual artifacts by utilizing cross-modal rewards. This method enhances spatial planning and aesthetic quality while ensuring verifiable correctness.

Key facts

Read aloud

AesCode introduces a novel approach to generating aesthetic code for visual artifacts by utilizing cross-modal rewards. This method enhances spatial planning and aesthetic quality while ensuring verifiable correctness. Generating information-rich visual artifacts such as slides, posters, and dashboards as code is appealing because the outputs are structured, editable, and verifiable. We therefore introduce AesCode, which generates code grounded in the content prompt while using an image generated from the same prompt as an aesthetic guidance. To evaluate this setting, we build a benchmark that scores verifiable correctness and visual quality separately, with visual scores highly consistent with human judgments. Relative to same-scale Qwen3-VL baselines given the same references, AesCode-8B and AesCode-32B improve visual quality by thirty one point one and twenty two point four percentage points.

Sources in this cluster