让神经网络按优化信号自动长出新路径，不提前预设结构
EMAN: Optimization-Driven Capacity Growth through Path Emergence in Multi-Task Learning
- 基于优化证据触发路径生长，而非固定架构或任务边界
- 仅在确认优化收益后才生成两条等容量独立路径
- 适合需要动态扩展能力的多任务学习场景
现有多任务学习方法依赖硬共享、多路径或专家模型、自适应共享和动态扩展，但其容量变化常受限于预设结构，或由任务边界与冲突信号触发。这引发一个根本问题：能否从单路径计算出发，仅在持续优化证据出现时才生长出独立新路径？我们提出涌现模块原子网络（EMAN），一种通过潜在相对相位暴露反称增长方向的优化驱动框架，无需即时实例化第二路径，并在训练中监控多个决策信号，将局部优化证据转化为结构决策。只有在验证通过后，才实现两条等容量独立路径的实体化。EMAN自适应分配共享与任务专属表征容量以适应不同任务需求。在控制秩设置、PASCAL-Context 和 NYUv2 上的大量实验验证了其有效性，在可比计算成本下取得更优性能。
原文摘要 · Abstract (English)
Existing multi-task learning methods rely on hard sharing, multiple paths or experts, adaptive sharing, and dynamic expansion. However, their capacity changes are usually constrained by predefined structures or triggered by task boundaries and conflict signals. This raises a fundamental question: can a network start from exact single-path computation and grow a new independent path only when persistent optimization evidence appears? We propose the Emergent Modular Atomic Network (EMAN), an optimization-driven framework for exposing an antisymmetric growth direction through latent relative phases without instantiating a second path, and for monitoring multiple decision signals during training to transform local optimization evidence into a structural decision. EMAN materializes two equal-capacity independent paths only after certification. EMAN adaptively allocates shared and task-specific representation capacity to accommodate varying task requirements. Extensive experiments on controlled rank settings, PASCAL-Context, and NYUv2 validate its effectiveness, achieving improved performance at a competitive computational cost.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。