NEMO: Execution-Aware Optimization Modeling via Autonomous Coding Agents 事件

NEMO: Execution-Aware Optimization Modeling via Autonomous Coding Agents arXiv:2601.21372v2 Announce Type: replace Abstract: We present NEMO, a system that translates Natural-language descriptions of decision problems into formal Executable Mathematical Optimization implementations using autonomous coding agents (ACAs). Existing approaches rely on specialized large language models (LLMs) or bespoke task-specific agents that are often brittle and frequently generate syntactically invalid or non-