Google 用 AI 智能体集群修复 TPU 编译器静默错误
Banger paper from Google on using agent swarms to fix Google's TPU compiler. The team used swarms o...
Google 让一群智能体去抓 TPU 编译器里最阴险的静默 bug，已经修了 142 个，还讲了人工审核的坑。
Google 团队发布了用智能体集群修复 TPU 编译器 bug 的论文。该框架已修复超过 142 个静默编译错误，另有 103 个发现正在审核中。静默错误不报错但会导致数值漂移或缓冲区损坏，可能毁掉一次大规模训练。方法上将合成图和生产图的 fuzzing 与 autotuner 并行检查结合，再用二分管线把故障定位到 fused kernel 内的单个操作，由智能体生成压力测试、复现故障并编写修复。
Banger paper from Google on using agent swarms to fix Google's TPU compiler. The team used swarms o...
Banger paper from Google on using agent swarms to fix Google's TPU compiler. The team used swarms of AI agents to find and fix silent compiler bugs, and the framework has resolved over 142 of them so far. They find silent miscompilations are the most dangerous. The compiler raises no error, but numbers drift, or buffers get corrupted, which can ruin a large training run or degrade live serving. The team combines fuzzing on synthetic and real production graphs with checks that run alongside the existing autotuner. A bisection pipeline narrows a failure in a full workload down to one operation inside a fused kernel. The agents generate stress tests, reproduce failures, and write fixes. The team steered them away from easy crash bugs and toward silent data corruption, and 103 more findings are under review. They also report that reviewer effort grew faster than the complexity of the fixes, and agent patches needed adversarial pre-filtering before humans looked at them. arxiv.org/abs/2611.00624 Chat with Paper: academy.dair.ai/papers/xla-fro… 💬 14 🔄 4 ❤️ 19 👀 1642 📊 16 ⚡
