Compiled AI: Deterministic Code Generation for LLM-Based Workflow Automation
编译AI:基于LLM的工作流自动化确定性代码生成
Geert Trooskens, Aaron Karlsberg, Anmol Sharma, Lamara De Brouwer, Max Van Puyvelde, Matthew Young, John Thickstun, Gil Alterovitz, Walter A. De Brouwer
机构
*
XY.AI Labs, Palo Alto, CA(XY.AI实验室,帕洛阿尔托)
;
Stanford University School of Medicine, Stanford, CA(斯坦福大学医学院,斯坦福)
;
Cornell University, Department of Computer Science, Ithaca, NY(康奈尔大学计算机科学系,伊萨卡)
;
Brigham and Women’s Hospital / Harvard Medical School, Boston, MA(布莱根妇女医院/哈佛医学院,波士顿)
Comments13 pages. Accepted at the 41st IEEE/ACM International Conference on Automated Software Engineering (ASE 2026). Artifact: https://doi.org/10.5281/zenodo.19245736
机构
*
Shanghai University of Finance and Economics(上海财经大学)
;
Meituan(美团)
;
The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳))
;
Peking University(北京大学)
CommentsWithdrawn at the request of ByteDance because the manuscript was submitted before completing the company's required internal review and approval process
SAT Certificates for the Matrix-Multiplication Challenges over F2: All Ten `Expected-UNSAT` Instances Are Satisfiable, and a Type-3-Free Rank-23 Scheme