Backend
On the Power of Alternating Local Certification with Constant-Size Certificates
arXiv cs.DC0 views
arXiv:2610.23694v1 Announce Type: new Abstract: We study alternation hierarchies for distributed decision in anonymous networks with constant-size certificates. A prover and a disprover alternately assign certificates to nodes over a constant number of rounds, after which each node decides whether to accept based on its own input and the certificates observed in its neighborhood. An instance is accepted if all nodes accept. Unlike previous work on distributed alternation hierarchies, which considered restricted local computation (Reiter, PODC 2024) or logarithmic-size certificates (Feuilloley et al., ICALP 2016), we focus on constant-size certificates and the role of neighborhood information in determining expressive power. We consider three models: SET, where nodes observe only the set of certificate sequences of their neighbors; MULTISET, where multiplicities are also visible; and PORTS, where neighbors can additionally be distinguished by local port numbers. We establish that SET is strictly weaker than MULTISET, which is in turn strictly weaker than PORTS, and that all three alternation hierarchies are strict at every level. Despite their severe restrictions, these models are remarkably expressive: LEADER belongs to the third level of SET. Our main characterization shows that, over finite input alphabets, SET coincides exactly with monadic second-order logic (MSO$_1$). Moreover, every language definable in counting monadic second-order logic (CMSO$_1$) belongs to PORTS, while MULTISET already contains languages that are not CMSO$_1$-definable. Nevertheless, this expressive power has fundamental limitations. PORTS does not capture MSO$_2$: neither Hamiltonicity nor the existence of a perfect matching belongs to PORTS. Furthermore, over unbounded input alphabets, PORTS cannot verify that every node has been given the correct network size.
Source:https://arxiv.org/abs/2610.23694PreviousBaGuaLu-2: Taming Dynamic Memory Fluctuation for Efficient MoE Model Training over 10,000 GPUsNextTRACE: Transformer Residual Activation Codebook Encoding for Communication-efficient Distributed Edge Inference
Related posts
find -size -1M Matched Only the Empty File: Size Units in a Camera Clip Cleanup Script
0dev.to
Mastering Background Tasks in Flutter: Efficient Data Sync & UI Updates
0dev.to
Remove a Photo Background in the Browser: An On-Device Model, No Upload
0dev.to
My morning terminal: zoxide jumps, mise runs the project, starship shows context, fzf picks the file
0dev.to

A stale translation passes every check and still tells a different story
0dev.to
NYT and publishers sue Brave over AI content scraping
0dev.to