
TL;DR
MoB introduces a game-theoretic expert routing mechanism using VCG auctions to improve continual learning by avoiding catastrophic forgetting and enabling emergent specialization without explicit task boundaries.
Contribution
It replaces learned gating networks with auction-based routing, providing stateless, incentive-compatible, and self-organizing expert selection in continual learning.
Findings
MoB achieves 88.77% accuracy on Split-MNIST, outperforming baselines.
Stateless routing in MoB prevents catastrophic forgetting.
Emergent specialization occurs without explicit task boundaries.
Abstract
Mixture of Experts (MoE) architectures have demonstrated remarkable success in scaling neural networks, yet their application to continual learning remains fundamentally limited by a critical vulnerability: the learned gating network itself suffers from catastrophic forgetting. We introduce Mixture of Bidders (MoB), a novel framework that reconceptualizes expert routing as a decentralized economic mechanism. MoB replaces learned gating networks with Vickrey-Clarke-Groves (VCG) auctions, where experts compete for each data batch by bidding their true cost -- a principled combination of execution cost (predicted loss) and forgetting cost (Elastic Weight Consolidation penalty). This game-theoretic approach provides three key advantages: (1) {stateless routing that is immune to catastrophic forgetting, (2) \textbf{truthful bidding} guaranteed by dominant-strategy incentive compatibility,…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsDomain Adaptation and Few-Shot Learning · Mobile Crowdsensing and Crowdsourcing · Stochastic Gradient Optimization Techniques
