# New Papers Push Language Agents Into Simulation Setup and Optimization Modeling

One system configures and self-corrects OpenFOAM computational fluid dynamics cases, another benchmark tests whether models can turn word problems into black-box optimization formulations.

- Published: 2026-08-05T05:32:18.568Z
- Canonical: https://polylog.news/ai/2026-08-05/new-papers-push-language-agents-into-simulation-setup-and-op
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [arXiv](https://arxiv.org/abs/2608.00003), [arXiv](https://arxiv.org/abs/2608.02612), [Meta AI](https://ai.meta.com/blog/genesis-mission-lawrence-berkeley-national-laboratory-segment-anything-dino/)

Two papers posted to arXiv today address the same constraint from different directions, which is the expert labor required to state an engineering problem in a form a solver can accept. AutoFOAM is described by its authors as a self-refinin…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-08-05/new-papers-push-language-agents-into-simulation-setup-and-op (subscription information: https://polylog.news/pricing).