Metraipolis › archive › dc_98a70edbb871
Does the Trust Boundary Hold over Time? A Long-Horizon Benchmark for Receiver-Directed Federated Learning — arXiv:2611.03724
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.03724
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-04T18:40:40.591Z
- SHA-256 of the archived text
2ea5d6706634606350fd1f667e00f0ac93e2d04c7923370bd444799ddf867c82- Stable URL of this page
https://metraipolis.com/documento/dc_98a70edbb871
How to cite
Does the Trust Boundary Hold over Time? A Long-Horizon Benchmark for Receiver-Directed Federated Learning — arXiv:2611.03724 — arXiv (en), https://arxiv.org/abs/2611.03724 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_98a70edbb871 (Metraipolis archive dc_98a70edbb871, sha256 2ea5d6706634606350fd1f667e00f0ac93e2d04c7923370bd444799ddf867c82)
BibTeX
@misc{dc98a70edbb871,
title = {Does the Trust Boundary Hold over Time? A Long-Horizon Benchmark for Receiver-Directed Federated Learning — arXiv:2611.03724},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.03724},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_98a70edbb871; sha256 2ea5d6706634606350fd1f667e00f0ac93e2d04c7923370bd444799ddf867c82},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03724. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
In receiver-directed federated learning, each site draws a trust boundary. For every source and model component it judges, on its own data, whether the transfer helps (admit), harms (reject) or cannot yet be told (abstain). We ask whether this boundary holds over hundreds of rounds. Our benchmark has a federation of thirteen semi-synthetic sites, an exact oracle for every evaluated transfer, paired seeds, an attack protocol, fourteen methods over 600 rounds and Qwen3-1.7B LoRA sites over 500. The answer depends on which boundary is measured. Across 20 paired CPU seeds, a gate judging against a fixed reference better identified transfers useful relative to the site's starting model. Yet it admitted 29% of transfers harmful to the model the site now has. A gate judging against the current model admitted 2% of harmful transfers on that target in the CPU tier, but missed many transfers useful to the starting model. Two short propositions explain these distinct failures. We also evaluate a held out audit that leaves gate verdicts intact. After training, it decides whether each final model is deployed or the site falls back to local training, and reports admissions that have gone stale. With independent audit data its guarantees hold after any training history. At 600 rounds the audit deployed 65 to 72% of models, none with negative benefit on its target, and kept 98% of the gain. No stale report had the wrong sign in our runs. With 20,000 audit observations, a variance-adaptive bound found 99% of stale transfers that cost at least 0.01 in log loss. A common-target Qwen3 comparison on 60 selected transfers shows the same tension over 100 rounds. With 38% adversarial sites, FedFomo and CS-Gate were the only evaluated methods with positive mean gain in every honest group. Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.DC, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Distributed, Parallel, and Cluster Computing (cs.DC). Pagina del lavoro: https://arxiv.org/abs/2611.03724 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-04T18:33:07.134Z dalla pagina https://arxiv.org/list/cs.DC/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).