提出MPCS框架，让模型持续学习时既记新知识又不忘旧知识。
MPCS: Neuroplastic Continual Learning via Multi-Component Plasticity and Topology-Aware EWC
- 融合11种机制，通过神经元动态生成与拓扑感知正则化实现自适应学习
- 在31项任务上综合表现优异，关键组件移除后性能下降超30个百分点
- 揭示了全局正则化反而损害性能，适合追求高效持续学习的研究者
持续学习系统面临可塑性（获取新知识）与稳定性（保留旧知识）的固有矛盾。本文提出MPCS（多可塑性持续学习系统），集成十一项互补机制：任务驱动神经发生、傅里叶编码输入、EWC正则化、元回放、混合巩固、混合门控、突触剪枝/再生、赫布更新、任务相似性路由、自适应生长控制及连续神经元重要性追踪。在覆盖回归、分类、逻辑与混合领域的31个任务的MEP-BENCH多轨道基准上，采用三维帕累托标准评估任务性能（Perf）、表示多样性（RD）和梯度冲突率（GCR）。15种消融配置（3种子×4轨迹×2000轮次）下，MPCS获得94.2的归一化效率得分，处于14个通过门槛系统中的帕累托前沿。关键发现：(i) 傅里叶编码为最关键组件（移除使性能下降30.7个百分点，14%任务未通过门槛）；(ii) 全局EWC降低性能（NES = -4.2）；拓扑局部EWC缓解此影响（90.5→91.8），但无法消除；完全移除EWC得MPCS_EFFICIENT，性能最高——在高任务相似性场景（s_bar ≈ 0.95）中呈现单调关系：全局EWC < 拓扑EWC < 无EWC；(iii) 帕累托状态评估具预测性：同时移除两个帕累托主导组件（EWC + 赫布）得MPCS_EFFICIENT，性能提升0.6个百分点，计算成本降低4.7倍（127分钟 vs. 602分钟），验证帕累托前沿可作为可行动的模型压缩指南。
原文摘要 · Abstract (English)
Continual learning systems face a fundamental tension between plasticity -- acquiring new knowledge -- and stability -- retaining prior knowledge. We introduce MPCS (Multi-Plasticity Continual System), a neuroplastic architecture that integrates eleven complementary mechanisms: task-driven neurogenesis, Fourier-encoded inputs, EWC regularization, meta-replay, mixed consolidation, hybrid gating, synapse pruning/regeneration, Hebbian updates, task similarity routing, adaptive growth control, and continuous neuron importance tracking. We evaluate MPCS on MEP-BENCH, a multi-track benchmark spanning 31 tasks across regression, classification, logic, and mixed domains, using a three-dimensional Pareto criterion over task performance (Perf), representation diversity (RD), and gradient conflict rate (GCR). Across 15 ablation configurations (3 seeds x 4 tracks x 2000 epochs), MPCS achieves a Normalized Efficiency Score of 94.2, placing it on the Pareto frontier among 9 of 14 gate-passing systems. Key findings: (i) Fourier encoding is the single most critical component (removal drops Perf by 30.7 pp and fails the MEP gate on 14% of tasks); (ii) global EWC degrades performance (NES = -4.2); topology-local EWC reduces this penalty (NES 90.5->91.8) but does not eliminate it; removing EWC entirely yields MPCS_EFFICIENT, the highest-Perf system -- establishing a monotone relationship in the high task-similarity regime (s_bar ~= 0.95): global EWC < topology EWC < no EWC; (iii) the Pareto status assessment is predictive: removing the two Pareto-dominated components (EWC + Hebbian) jointly yields MPCS_EFFICIENT, which improves Perf by 0.6 pp at 4.7x lower compute cost (127 vs. 602 min), validating the Pareto frontier as an actionable model-compression guide.
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。