Google DeepMind 与 MIT 等发布 SMART 论文:设计文档取代代码作为 ML 性能建模库的核心产物

elvis · @omarsar0 · X·2026-09-07 23:25·22分钟前
AI 导读

Google DeepMind、MIT、Stanford 等机构提出 SMART,一个 ML 性能建模库,其 main 分支几乎不含代码,仓库是自然语言设计文档的有向图,编码子智能体在版本更新时从文档重新生成全部实现。

elvis@omarsar0
57AI 编辑部评分,满分 100

Google DeepMind 与 MIT 等发布 SMART 论文:设计文档取代代码作为 ML 性能建模库的核心产物

2026-09-07 23:25· 22分钟前
AI 导读

Google DeepMind、MIT、Stanford 等机构提出 SMART,一个 ML 性能建模库,其 main 分支几乎不含代码,仓库是自然语言设计文档的有向图,编码子智能体在版本更新时从文档重新生成全部实现。

// Design Docs Are All You Need //

Banger paper from Google DeepMind, MIT, and colleagues.

What a genuinely strange and interesting paper this one is.

Here is the setup:

They maintain a performance-modeling library whose main branch contains almost no code.

The repository is a directed graph of natural-language design docs. Coding sub-agents regenerate the entire implementation from those docs whenever a version updates.

Every human change is an edit to a doc.

The premise is that ML performance modeling invalidates its own abstractions every hardware and model generation, and coding agents are now cheap enough that regenerating a library beats patching one.

Two things make the regeneration reliable. The design docs are written around step-by-step worked examples, which act as in-context demonstrations for the generating agents. The system is also anchored on a minimal recursively defined operator IR with symbolic cost expressions in SymPy.

Regenerated implementations reproduce hand-audited reference models to round-off precision, including DeepSeek-V3 serving on a TPU pod slice.

Paper: https://academy.dair.ai/papers/design-docs-are-all-you-need-an-ai-native-machine-learning-performance-tool-2609.05364