外观
论文arXiv 2610.2227613 天前
MJev-Spatial：通过原生注意力匹配实现 System One 空间智能
原文链接：MJev-Spatial: System One Spatial Intelligence via Native Attention Matching（arXiv 2610.22276 · 发布日期：2026-10-15）
MJev-Spatial: System One Spatial Intelligence via Native Attention Matching
MJev-Spatial面向交互速率空间判断，将选择、数值和点预测统一为分类，并在空间问答与12个基准上验证。
核心要点
- VLM以自由文本回答紧凑空间问题，需自回归解码和答案解析，影响交互速率。
- 模型把候选项、数值桶和图像区域视为类别，用预训练查询/键投影打分且不增参数。
- 推理单次前向且不生成答案词元；联合监督粗到细阶段，训练吞吐提升30.8%。
- 基于Qwen3.5-4B和12.3M问答训练，在八个模型中总分最高并快于自回归基线。
验证范围与局限：验证限于由现有空间数据集转换的问答训练和12个基准测试。
英文摘要 (Abstract)
Spatial intelligence lets an agent perceive and act in the physical world, and many of its judgments must be made at interaction rates: which spatial relation holds, how far an object is, or where to place it. Vision-language models (VLMs) produce these answers by free-form text generation, paying for autoregressive decoding and answer parsing even though the output space is compact: one of a few options, a scalar, or a 2D point. In light of this, we introduce MJev-Spatial, a System 1 multimodal decision model that tackles all three types of spatial understanding tasks within a unified classification framework. Specifically, MJev-Spatial treats each candidate as a class and scores it using the backbone's pretrained query and key projections, without adding parameters. The candidates are the given options for choice tasks, or value bins and image regions refined through coarse-to-fine classification for numerical and point prediction, respectively. At inference, each classification decision takes a single forward pass and generates no answer tokens. To train on heterogeneous spatial data efficiently, we supervise all coarse-to-fine stages jointly with teacher forcing in one forward pass, and combine this with length-aware global batching and cross-GPU load balancing, raising training throughput by 30.8% over the original batching strategy. Built on Qwen3.5-4B and trained on 12.3M question-answer pairs converted from existing spatial datasets, MJev-Spatial achieves the highest overall score (65.64%) among eight tested models, ranks first on seven of twelve benchmarks, and outperforms its unmodified autoregressive base model on all twelve, while running 2.5 times faster than the fastest autoregressive baseline and 5.6 times faster than the base model.
作者：Luozheng Qin, Yuhan Zhu, Changhai Zhou, Qian Qiao, Jun Gao, Yuhua Zhou, Harry Zhang, Xiangyu Zeng, Yuxuan Liu, Kaijie Chen
来源
本页事实对应的公开页面，以原文为准。