Scalable, high-capacity, and low-power computing architecture is the primary assurance for increasingly manifold and large-scale machine learning tasks. Traditional electronic artificial agents by conventional power-hungry processors have faced the issues of energy and scaling walls, hindering them from the sustainable performance improvement and iterative multi-task learning. Referring to another modality of light, photonic computing has been progressively applied in high-efficient neuromorphic systems. Here, we innovate a reconfigurable lifelong-learning optical neural network (L²ONN), for highly-integrated tens-of-task machine intelligence with elaborated algorithm-hardware co-design. Benefiting from the inherent sparsity and parallelism in massive photonic connections, L²ONN learns each single task by adaptively activating sparse photonic neuron connections in the coherent light field, while incrementally acquiring expertise on various tasks by gradually enlarging the activation. The multi-task optical features are parallelly processed by multi-spectrum representations allocated with different wavelengths. Extensive evaluations on free-space and on-chip architectures confirm that for the first time, L²ONN avoided the catastrophic forgetting issue of photonic computing, owning versatile skills on challenging tens-of-tasks (vision classification, voice recognition, medical diagnosis, etc.) with a single model. Particularly, L²ONN achieves more than an order of magnitude higher efficiency than the representative electronic artificial neural networks, and 14× larger capacity than existing optical neural networks while maintaining competitive performance on each individual task. The proposed photonic neuromorphic architecture points out a new form of lifelong learning scheme, permitting terminal/edge AI systems with light-speed efficiency and unprecedented scalability.

可扩展、大容量、低功耗的计算架构是日益多样化、大规模的机器学习任务的首要保证。由传统的耗电处理器组成的传统电子人工智能体面临着能源和缩放壁的问题，阻碍了它们可持续的性能改进和迭代多任务学习。参考光的另一种形态，光子计算已逐渐应用于高效的神经形态系统。在这里，我们创新了一种可重构的终身学习光学神经网络（L ² ONN），通过精心设计的算法-硬件协同设计来实现高度集成的数十任务机器智能。受益于大规模光子连接固有的稀疏性和并行性，L ² ONN 通过自适应激活相干光场中的稀疏光子神经元连接来学习每个单一任务，同时通过逐渐扩大激活来逐步获取各种任务的专业知识。多任务光学特征通过分配不同波长的多光谱表示并行处理。^{对自由空间和片上架构的广泛评估证实，L 2} ONN首次避免了光子计算的灾难性遗忘问题，拥有应对数十种任务（视觉分类、语音识别、医疗诊断、等）与单一模型。特别是，L ² ONN 的效率比代表性的电子人工神经网络高一个数量级以上，容量比现有的光学神经网络大 14 倍，同时在每个单独任务上保持竞争性能。所提出的光子神经形态架构指出了一种新形式的终身学习方案，允许终端/边缘人工智能系统具有光速效率和前所未有的可扩展性。