跳到正文
arXiv cs.AI· Tobias Ladner, Konstantin Kaulen, Stanley Bak, Julien Girard-Satabin, Taylor T. Johnson, Edoardo Manino, ThanhVu H Nguyen, Ben Wooding, Haoze Wu·· 10 小时前AI 评分18

VNN-COMP 2026 神经网络验证竞赛总结与结果

The 7th International Verification of Neural Networks Competition (VNN-COMP 2026): Summary and Results

AI 导读

第七届国际神经网络验证竞赛(VNN-COMP 2026)在葡萄牙里斯本 FLoC 2026 的 CAV 会议期间举办,报告总结了赛事规则、基准、参赛工具与结果。竞赛统一采用 ONNX 表示神经网络、VNNLIB 表示规约,所有工具由组织者在 AWS 上运行,以可扩展性和速度客观比较当前神经网络验证方法。

正文 · 原文

View PDF HTML (experimental)

Abstract:This report summarizes the seventh International Verification of Neural Networks Com- petition (VNN-COMP 2026), held as part of the Symposium on AI Verification (SAIV) co-located with the International Conference on Computer-Aided Verification (CAV) at FLoC 2026 in Lisbon, Portugal. The goal of the competition is to provide an objective comparison of the state-of-the-art methods in neural network verification, in terms of scal- ability and speed. Along this line, we used standard formats (ONNX for neural networks and VNNLIB for specifications), standard hardware (all tools are run by the organizers on AWS), and tool parameters provided by the tool authors. This report summarizes the rules, benchmarks, participating tools, results, and lessons learned from this competition.
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.24114 [cs.LG]
  (or arXiv:2611.24114v1 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2611.24114

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Julien Henri-Aurélien Girard-Satabin [view email]
[v1] Wed, 18 Nov 2026 09:30:13 UTC (2,021 KB)

来源:arXiv cs.AI · arxiv.org