Design Docs Are All You Need: An AI-native Machine-Learning Performance Tool

• URL: http://arxiv.org/abs/2609.05364v1
• Date: Fri, 04 Sep 2026 17:08:51 GMT
• Title: Design Docs Are All You Need: An AI-native Machine-Learning Performance Tool
• Abstract summary: Machine-learning performance modeling is a uniquely hostile terrain for long-lived software.<n>We describe a rigorous symbolic performance-modeling library for ML systems whose main branch contains almost no code.<n>Every human change is a natural-language edit to a doc--self-documenting by construction.
• Score: 11.868531447858372
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Machine-learning performance modeling is a uniquely hostile terrain for long-lived software: the assumptions baked into today's abstractions are invalidated by tomorrow's models and systems, forcing perpetual refactoring of performance-modeling frameworks. Meanwhile, AI coding agents have become fast and capable enough that regenerating an entire library is cheaper than paying down the tech debt of incrementally patching it. We describe SMART, a rigorous symbolic performance-modeling library for ML systems whose main branch contains almost no code: the repository is a DAG of self-contained natural-language design docs, coding sub-agents regenerate the implementation from only the docs on new version updates, and every human change is a natural-language edit to a doc--self-documenting by construction. Two ingredients make regeneration reliable: (i) a design-doc style built around step-by-step worked examples that act as in-context demonstrations for the generating agents, and (ii) a minimal, recursively defined operator IR with symbolic (SymPy) cost expressions, a fast analytical roll-up mode for large sweeps, and a slow modulo-scheduling mode for fine-grained schedule studies. Regenerated implementations reproduce hand-audited reference models--including DeepSeek-V3 serving on a TPU pod slice--to round-off precision, suggesting that design docs--not code--can be the durable artifact for ML-systems co-design tools.

Related papers

• Code as Representation: A Compilable Parsing Paradigm for Academic Documents [50.40789738182074]
  Compilable Academic Document Parsing (CADP) is a paradigm that reconstructs a full page as contextual La plus executable Python.<n>We introduce CADP-Bench, an expert-verified benchmark of full academic pages containing text and multiple SAE types.<n>Results show that even frontier models still struggle to produce high-fidelity executable reconstructions.
  arXiv  Detail & Related papers  (2026-08-18T09:10:43Z)
• Model-Driven Discipline for Multi-Agent LLMs: Requirement-to-Verification Generation of Traceable System Models [62.62160094849002]
  RADIANT is an engineering methodology that combines MDE with Multi-Agent Large Language Models.<n>It automatically generates heterogeneous models across engineering phases.<n>It provides exact, automated change-impact analysis.
  arXiv  Detail & Related papers  (2026-07-18T08:50:55Z)
• The Illusion of Agentic Complexity in README.md Generation: Evaluating Single-Agent vs. Multi-Agent RAG Systems [3.7149543017244238]
  This study empirically evaluates Retrieval-Augmented Generation (RAG) dependent architectures for the generation of files for GitHub repositories.<n>Results indicate a critical architectural trade-off: the Single-Agent pipeline achieves lexical quality comparable to MAS while reducing token consumption by 86% and operating at twice the speed.
  arXiv  Detail & Related papers  (2026-06-29T16:29:31Z)
• Understanding by Reconstruction: Reversing the Software Development Process for LLM Pretraining [66.89012795621349]
  Large Language Models (LLMs) often struggle with the deep, long-horizon reasoning required for complex software engineering.<n>We propose a novel paradigm: understanding via reconstruction.<n>We introduce a framework that synthesizes latent agentic trajectories using a multi-agent simulation.
  arXiv  Detail & Related papers  (2026-03-11T09:23:20Z)
• Architecture-Aware Multi-Design Generation for Repository-Level Feature Addition [53.50448142467294]
  RAIM is a multi-design and architecture-aware framework for repository-level feature addition.<n>It shifts away from linear patching by generating multiple diverse implementation designs.<n>Experiments on the NoCode-bench Verified dataset demonstrate that RAIM establishes a new state-of-the-art performance.
  arXiv  Detail & Related papers  (2026-03-02T12:50:40Z)
• Steerable Instruction Following Coding Data Synthesis with Actor-Parametric Schema Co-Evolution [30.919665619254]
  We propose IFCodeEvolve, a framework for synthesizing large-scale instruction-paired coding data.<n>We show that IFCodeEvolve significantly boosts base model performance, with our 32B model achieving parity with proprietary SOTA models.<n>We also contribute IFCodeBench, a comprehensive human-verified benchmark equipped with solutions and robust AST-based verification.
  arXiv  Detail & Related papers  (2026-02-27T21:31:41Z)
• Generating Software Architecture Description from Source Code using Reverse Engineering and Large Language Model [2.6126272668390373]
  Software Architecture Descriptions (SADs) are essential for managing the inherent complexity of modern software systems.<n>SADs are often missing, outdated, or poorly aligned with the system's actual implementation.<n>We propose a semi-automated generation of SADs from source code by integrating reverse engineering (RE) techniques with a Large Language Model (LLM)
  arXiv  Detail & Related papers  (2025-11-07T11:35:46Z)
• Every Step Counts: Decoding Trajectories as Authorship Fingerprints of dLLMs [63.82840470917859]
  We show that the decoding mechanism of dLLMs can be used as a powerful tool for model attribution.<n>We propose a novel information extraction scheme called the Directed Decoding Map (DDM), which captures structural relationships between decoding steps and better reveals model-specific behaviors.
  arXiv  Detail & Related papers  (2025-10-02T06:25:10Z)
• Repeton: Structured Bug Repair with ReAct-Guided Patch-and-Test Cycles [1.387448620257867]
  Large Language Models (LLMs) have shown strong capabilities in code generation and comprehension, yet their application to complex software engineering tasks often suffers from low precision and limited interpretability.<n>We present Repeton, a fully open-source framework that leverages LLMs for precise and automated code manipulation in real-world Git.
  arXiv  Detail & Related papers  (2025-06-09T19:36:40Z)
• Paper2Code: Automating Code Generation from Scientific Papers in Machine Learning [70.04746094652653]
  We introduce PaperCoder, a framework that transforms machine learning papers into functional code repositories.<n>PaperCoder operates in three stages: planning, designs the system architecture with diagrams, identifies file dependencies, and generates configuration files.<n>We then evaluate PaperCoder on generating code implementations from machine learning papers based on both model-based and human evaluations.
  arXiv  Detail & Related papers  (2025-04-24T01:57:01Z)
• VeriMind: Agentic LLM for Automated Verilog Generation with a Novel Evaluation Metric [4.590930025882158]
  We propose VeriMind, an agentic LLM framework for Verilog code generation.<n>We introduce a novel evaluation metric-pass@ARC-which combines the conventional pass@k measure with Average Refinement Cycles (ARC) to capture both success rate and the efficiency of iterative refinement.<n> Experimental results on diverse hardware design tasks demonstrated that our approach achieved up to $8.3%$ improvement on pass@k metric and $8.1%$ on pass@ARC metric.
  arXiv  Detail & Related papers  (2025-03-15T23:43:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.