# Praxist：从实验性产物到解决方案谱系

> 原文链接：https://robotics.hellobike.com/papers/2026-08-26-2608-25955
> 频道：papers | 更新日期：2026-08-26 | 标签：arxiv, cs.RO
> 核心摘要：Praxist用谱系中心的证据图与代际系统解决自主R&D反复重学问题，并在MLE-bench和五个工程案例验证。

# Praxist：从实验性产物到解决方案谱系

::: v-pre
> 原文链接：[Praxist: From Experimental Artifacts to Solution Lineages](https://arxiv.org/abs/2608.25955)（arXiv 2608.25955 · 发布日期：2026-08-26）

**Praxist: From Experimental Artifacts to Solution Lineages**

Praxist用谱系中心的证据图与代际系统解决自主R&D反复重学问题，并在MLE-bench和五个工程案例验证。

## 核心要点

- 现有自主R&D系统将尝试近似自包含，难追踪改进来源并反复重学经验。
- Praxist把可复现产物和评估结果转为类型化证据图、分道前沿与议程。
- 在75任务MLE-bench上获60枚奖牌、49金，成本低于Claude Code基线。
- 五个工程案例中相对各自任务原生基线提升准确性、生存率或资源成本。

**验证范围与局限**：验证范围为标准化75任务MLE-bench套件及五个开放式工程问题案例。

## 英文摘要 (Abstract)

Autonomous R&D agents now write, run, and improve executable artifacts under automated evaluation-but largely as laboratory instruments: shown on curated benchmarks, with gains that are hard to trace to a cause and costs well above what sustained engineering practice absorbs. The limitation is structural. Most systems treat each attempt as nearly self-contained, so logs, memories, and search trees record what happened without establishing which design element produced an improvement, whether its evidence survived validation, or how it recombines with others. Long campaigns therefore keep re-learning the same lessons. We introduce Praxist, a lineage-centered generational system that converts reproducible artifacts and evaluator outcomes into a typed evidence graph of findings, lane-structured frontiers, and agendas. Separating local artifact construction from cohort-level evidence synthesis lets later attempts inherit validated mechanisms, unresolved claims, and useful constraints, and leaves results attached to an inspectable lineage. On the standardized 75-task MLE-bench suite, the finalized official-grader results give Praxist 60 medals (80.0%), 49 of them gold, against 55 medals (73.3%) and 34 gold for a Claude Code baseline on Claude Opus 4.8-at a recorded model spend of $3054 versus $38370, roughly a twelfth of the cost. Five case studies-quantitative trading, LiDAR-inertial-visual SLAM, tokamak magnetic control, rocket landing, and job-shop rescheduling-carry the same process into open-ended engineering problems, improving on each task-native baseline in headline accuracy, survival, or resource cost, with the discovery path on record. Stronger artifacts at an order of magnitude less spend, each backed by an auditable lineage, are, to our knowledge, first brought together here: the operating profile production research requires, not the one a benchmark demonstration establishes.

作者：Jin Li, Ahmed Murtadha, Zhiyu Wang, Qiwen Chen, William C. Chen, Yifei Wu, Guan Wang, Cheng Gong, Andy L. Siy, Jiayi Yang
:::

## 原始参考来源 (Sources)

- [arXiv 2608.25955：Praxist: From Experimental Artifacts to Solution Lineages](https://arxiv.org/abs/2608.25955)