Designer-RSI：冻结模型靠程序记忆把设计成功率提到99.3%
Designer-RSI: Evolving Procedural Memory from User Traffic for Agentic Graphic Design
不微调不标注，让 Claude 靠一份会自我修订的技能清单，把设计成功率从72.7%刷到99.3%。
论文提出 Designer-RSI 框架：冻结的 Claude-Sonnet-4 通过 230 多个工具操作专业设计软件，外部程序记忆以自然语言技能形式积累可复用的设计流程。记忆通过加宽机制为新子任务补充流程、通过加深机制依据成功与失败执行修订既有流程，重放门槛只保留修好失败且不损害成功结果的改动。在 1,406 条真实用户需求上跑 5 轮、共 1,869 条自动评分轨迹，技能库从 76 条扩到 139 条，全程无权重更新、无人工标注。GenEval2 执行成功率从 72.7% 升至 99.3%，生成质量提升 11.99 分；在四个专业设计基准上，对无技能基线分别取得 61.8%（Claude-Sonnet-4）和 67.6%（Claude-Opus-4.6）胜率。消融显示两种机制叠加有效：200 条保留需求上单用加宽或加深胜率为 49.4%/48.6%，组合达 58.5%（p=0.025）。
Designer-RSI: Evolving Procedural Memory from User Traffic for Agentic Graphic Design
Professional graphic design is a long-horizon agentic task in which structured, editable artifacts emerge from many interdependent actions, yet outcomes admit no reliable programmatic oracle. We introduce a continual adaptation framework in which a frozen frontier model operates professional design software through more than 230 tools, while an external procedural memory of natural-language skills accumulates and refines reusable design procedures from experience. The memory widens by acquiring procedures for recurring uncovered subtasks and deepens by revising existing procedures against their own successful and failed executions, while a matched replay gate admits only changes that repair failures without regressing observed successes. Five rounds over 1,406 real user briefs and 1,869 automatically graded trajectories, with no weight updates and no human labels, grow the bank from 76 documentation-derived skills to 139 and raise GenEval2 execution success on Claude-Sonnet-4 from 72.7% to 99.3% (+11.99 points in generation quality), with 61.8% and 67.6% win rates against the no-skill agent across four specialized design benchmarks on Claude-Sonnet-4 and Claude-Opus-4.6. We further show the two mechanisms are effective in combination: on 200 held-out briefs from user-traffic benchmark, widening or deepening alone reaches a 49.4% / 48.6% win rate over the no-skill agent, while their combination reaches 58.5% (p = 0.025). Procedural memory offers a practical route to continual adaptation of agents under noisy, unverifiable feedback.
