iARCS: פרקטיקה רפורמטורית לפיתוח סביבתי סימולטורית 3D

תקציר מקורי באנגליתarXiv:2608.06161v3 Announce Type: replace Abstract: Synthetic 3D scene generation is increasingly used as a data source for computer vision and embodied AI, but existing generators often optimize perceptual realism without reliably satisfying task-critical functional constraints. This mismatch limits the usefulness of synthetic data for downstream training, where accessibility, traversability, and spatial rule compliance are often essential. We present iARCS, an iterative agentic reinforcement learning framework that adapts a pretrained scene generator to naturallanguage task requirements. iARCS uses a two-phase strategy: universal-reward pretraining to improve physical plausibility and layout quality, followed by task-specific finetuning with LLM-generated reward programs that are iterati