arXiv cs.AI· Weilin Ruan, Bohan Zhou, Jinghao Li, Yangkai Wei, Wenqian Li, Peiran Dong, Yingjie Xu, Leyi Wu, Haojian Huang, Binghui Xie, Han Yang, Kaiwen Zhou, Yinchuan Li, Ying-Cong Chen, James Cheng·· 2 天前AI 评分29
Anvil：通过知识再循环扩展具身空间智能
Anvil: Scaling Embodied Spatial Intelligence through Knowledge Recirculation
AI 导读
研究团队提出 Anvil 统一框架，用于具身空间智能的数据生成、训练与评测，包含数据飞轮 Anvil-Gen、场景对齐监督 Anvil-Ben 和 9B 蒸馏学生模型 Anvil-MOPD。
正文 · 原文
Authors:Weilin Ruan, Bohan Zhou, Jinghao Li, Yangkai Wei, Wenqian Li, Peiran Dong, Yingjie Xu, Leyi Wu, Haojian Huang, Binghui Xie, Han Yang, Kaiwen Zhou, Yinchuan Li, Ying-Cong Chen, James Cheng
Abstract:Embodied spatial intelligence requires more than visual recognition: models must localize entities, reason about their relations, and express this knowledge in action-relevant forms. Yet joint training can conflate the knowledge needed at different stages because supervision is fragmented across incompatible geometric and semantic outputs. We introduce Anvil, a unified framework for data generation, training, and benchmarking for embodied spatial intelligence. (1) Anvil-Gen is a heterogeneous data flywheel that combines form-specific generation, cross-stage consistency verification, and failure-driven recollection across 11 Ground--Relation--Act forms. (2) Anvil-Ben provides scene-aligned supervision for learning diverse embodied spatial capabilities and scene-linked assessments that distinguish stage competence from error propagation. (3) Anvil-MOPD is a 9B student that consolidates heterogeneous spatial knowledge through distillation from specialist teachers. Semantic-first-fault routing targets teacher supervision to the student's earliest decisive failure, while stage normalization controls the loss scale across native outputs of different lengths. We construct 5.8M source-grounded records spanning 160K images, reserving 580K spatial records for held-out evaluation and using the remaining 5.22M records for training. On average across seven public benchmarks, Anvil-MOPD outperforms Qwen3.5-9B and RynnBrain-8B by 9.29 and 11.03 percentage points, respectively, while reaching 80.67% on VABench-Point. Controlled studies support targeted coverage, selective transfer, and feedback-guided repair as complementary framework components.
| Comments: | 28 pages, 7 figures |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.21882 [cs.CV] |
| (or arXiv:2610.21882v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.21882 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Weilin Ruan [view email]
[v1]
Thu, 15 Oct 2026 13:47:08 UTC (4,294 KB)
来源：arXiv cs.AI · arxiv.org