UCSD Hao AI Lab 与 SGLang 社区合作,将 vstack 用于 SGLang 的智能体形式化验证,补充常规单元与 e2e CI,通过穷举异步交错与状态转换来捕获并发和多步 bug。团队称已提出 20+ 个 issue 与修复 PR,推文中分享了其中 5 个,涵盖权重更新竞态条件到 decode server 卡顿。vstack 已在 GitHub 开源。
(1/8) Agentic Formal Verification on SGLang 🚀
Last week, we introduced vstack for agentic formal verification in AI infra; now we're working with the @sgl_project community on its mature, robust serving stack.
vstack complements regular unit and e2e CI by comprehensively exploring async interleavings and state transitions, catching concurrency and multi-step bugs before they hit production. Below we share 5 of the 20+ issues and fix PRs we proposed with vstack, from weight-update race conditions to decode server stalls🧵
🔌 Try vstack here: github.com/hao-ai-lab/vstack
来源:Hao AI Lab · x.com