arXiv：cs.LG· Junlin Chen, Wenhao Wang, Junchao Xu, Qingsong Yao, Fei Xue, Li Sun, Xiao Wang, Ruijie Wang·· 10 小时前
MAEM：面向类增量图学习的无回放、容量可控动态专家混合框架，缓解近因偏差
MAEM: Mitigating Recency Bias via a Replay-Free, Capacity-Controlled Dynamic Mixture of Experts for Class-Incremental Graph Learning
AI 导读
MAEM 是一个无回放、容量可控的动态专家混合框架，用于类增量图学习。它用 SGC 构建表示并为每个新会话训练独立专家后冻结，各专家以 MAE 重构误差最低者完成无任务 ID 路由；专家池达到固定容量后，MAEM 通过 Git Re-Basin 对齐合并最少使用的专家，并用 MAE 分布合成的拓扑感知伪数据将其蒸馏为融合专家。
正文
Abstract:Graph Neural Networks (GNNs) excel on graph-structured data but suffer catastrophic forgetting when classes arrive sequentially. In class-incremental graph learning, predictions are often dominated by recently learned classes, causing severe recency bias. Dynamic expansion can isolate knowledge across sessions, but many methods assume task-incremental settings and require task identities at test time to select branches or heads---information unavailable in class-incremental learning. Learnable routers offer an alternative, yet continual fine-tuning the router on new data can bias routing toward recent components. We propose MAEM, a replay-free, capacity-controlled dynamic mixture-of-experts framework for class-incremental graph learning. MAEM builds representations with Simplified Graph Convolution (SGC) and trains a dedicated expert for each new session, then freezes it to protect prior knowledge. For routing without task IDs or a shared trainable router, each expert uses a masked autoencoder (MAE); test nodes are assigned to the expert with the lowest reconstruction error, reducing routing-level recency bias. When the expert pool reaches a fixed capacity, MAEM merges the least-used experts via Git Re-Basin alignment and further distills them into a fused expert using topology-aware pseudo data synthesized from MAE distributions (DeepInversion-style), without storing or revisiting historical graph data. On four class-incremental graph benchmarks, MAEM mitigates recency bias, outperforms strong continual graph learning baselines, and keeps model capacity bounded under strict replay-free learning.
| Subjects: | Machine Learning (cs.LG); Machine Learning (stat.ML) |
| Cite as: | arXiv:2610.20238 [cs.LG] |
| (or arXiv:2610.20238v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.20238 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Junlin Chen [view email]
[v1]
Wed, 14 Oct 2026 14:16:40 UTC (260 KB)
来源：arXiv：cs.LG · arxiv.org