AutoDesign：長時程智能體設計的後設框架優化

摘要

將多模態來源轉化為精簡且結構化的媒體輸出，從根本上可被概念化為一個以模型-框架系統為核心的長視野代理過程。理想的框架系統理應與人類設計先驗對齊，並透過經驗探索累積可重複使用的經驗，以驅動遞迴自我改進；然而，現有範式依然靜態，未能達成此能力。在本文中，我們提出AutoDesign——一個與人類設計先驗對齊的架構，其中元框架優化器引導程式碼代理根據展開回饋遞迴地改進框架。為實例化並評估此架構，我們聚焦於學術論文到海報生成任務，並引入PosterBench，其包含涵蓋五個學科的100篇論文主軌道（Main Track），以及用於受控評估的共享10篇論文子集PosterBench-mini。在PosterBench Main Track上，AutoDesign取得最高分78.32，以7.45分的優勢超越封閉原始碼商業系統Claude Design。在七種受控的程式碼代理-模型配置中，整合所學到的DesignHarness持續提升效能，將平均PosterBench分數從54.99提升至67.39（+12.4%）。在完全自主的長視野迴圈中，AutoDesign在40分鐘內以不到3美元的成本執行253次工具呼叫與11次編輯回合，於人工評估中達到平均會議海報品質。一項系統盲測人工研究進一步顯示，在受評系統中，AutoDesign獲得最高的人類偏好。

English

Transforming multimodal sources into condensed and structured media outputs can be fundamentally conceptualized as a long-horizon agentic process centered on a model-harness system. While an ideal harness system should align with human design priors and accumulate reusable experience through empirical exploration to drive recursive self-improvement, existing paradigms remain static and fall short of this capability. In this paper, we present AutoDesign, a framework that aligns with human design priors, where a meta-harness optimizer guides a code agent to recursively improve harness based on rollout feedback. To instantiate and evaluate this framework, we focus on the academic paper-to-poster generation task and introduce PosterBench, comprising a 100-paper Main Track spanning five disciplines and PosterBench-mini, a shared 10-paper subset for controlled evaluation. On the PosterBench Main Track, AutoDesign achieves the highest score of 78.32, surpassing the closed-source commercial system Claude Design by 7.45 points. Across seven controlled code-agent-model configurations, integrating the learned DesignHarness consistently improves performance, increasing the average PosterBench Score from 54.99 to 67.39 (+12.4%). In a fully autonomous long-horizon loop, it executes 253 tool calls and 11 editing turns within 40 minutes for under $3, reaching average conference-poster quality in human evaluation. A system-blind human study further demonstrates that AutoDesign achieves the highest human preference among evaluated systems.