With the development of sharded blockchains, account migration mechanisms migrate accounts selected by account partition algorithms from the source shard to the target shard, aiming to reduce cross-shard transactions and balance the load. Moreover, the cost of account migration plays a critical role in determining the effectiveness of partitioning algorithms, as performance gains from resharding depend on cost-efficiency. Therefore, an efficient and low-overhead migration mechanism is critical to improving the performance of sharded blockchains. Existing migration approaches require cross-shard locks or intra-shard locks, which causes service interruption. However, we observe that account states remain unchanged during migration, indicating that such lock-based mechanisms are overly restrictive. To this end, we propose a novel Lock-Free Account Migration ledger, denoted as LFAMLedger, which consists of a prewrite phase and a migration phase. LFAMLedger ensures data consistency and lock-free migration through timestamp-based multi-version snapshots and a deferred validation mechanism. Furthermore, we propose optimistic snapshot isolation to decouple the source and target shards, and a flexible strategy to enable concurrent migration of accounts and their associated transactions to multiple target shards. We conduct extensive scalability experiments based on real Ethereum transaction histories. Compared with cross-shard lock-based and intra-shard lock-based migration approaches, our lock-free LFAMLedger achieves 1.9× and 1.8× higher throughput, and reduces latency by 2.35× and 1.75×, respectively, demonstrating significant improvements in efficiency.
Shuai Zhao, Zhiwei Zhang, Jun-Kai Wang et al.· IEEE Transactions on Knowled...· 0 citations
Kernel methods, which embed data distributions into a reproducing kernel hilbert space (RKHS) via positive-definite similarity measures, continue to play an important role. However, learning a good, generalizable kernel for high-dimensional and heterogeneous data under temporal or regional distribution shift remains challenging. To address these issues, we propose SpectraMancer, which learns kernels directly in the Fourier spectral domain induced by multilevel circulant matrices, thereby enabling generalizable kernel learning for complex data. SpectraMancer embeds all shift-invariant candidates into a common multilevel order via randomized multilevel circulant matrices, which yields a fixed Fourier diagonalization and turns inverses, products, and gradients into elementwise fast Fourier transform (FFT) operations. To the best of our knowledge, this is the first kernel-learning approach that exploits randomized multilevel circulant matrices for joint diagonalization across kernels. SpectraMancer further enforces scale invariance via kernel double centering and Frobenius normalization, reduces spectral variance through antithetic phase pairing with quasi-Monte Carlo draws, and optimizes a solver-free spectral risk proxy (SRP) for bandwidth weighting without repeated inner solves. Experimental results show that SpectraMancer improves spectrum-aware kernel selection and predictive performance across diverse benchmarks.
Li-Zhong Ding, Jiarun Fu, Qiuning Wei et al.· IEEE Transactions on Neural...· 0 citations
Persistent Consistency Self-Distillation (PCSD) is proposed, which derives token-level distillation weights from the local persistence of teacher-favoring signals, and combines adaptive windows with exponentially decayed aggregation to capture persistent relative teacher support.
Chunji Lv, Yangguang Wei, Junlin Liu et al.· 0 citations
FedGAMMA is proposed, casting federated multimodal graph foundation learning as a two-stage semantic-structural alignment problem of federated pre-training and prompt-based fine-tuning, and outperforms competitive baselines accross multi-domain datasets on multiple tasks.
Xunkai Li, Guohao Fu, Yuming Ai et al.· 0 citations
CHILL-Harness intervenes at the orchestration layer to enable advantage-guided workflow adaptation, thereby improving reasoning and execution efficiency while preserving task performance and incorporating a success-preserving objective and advantage-margin authorization constraints into CHILL-Harness to promote reliable adaptation.
