Designer-RSI: 사용자 트래픽으로부터 에이전트형 그래픽 디자인을 위한 절차적 메모리 진화시키기
이 논문은 사용자의 트래픽으로부터 procedural memory를 진화시키는 Designer-RSI 프레임워크를 제안합니다. frozen frontier model이 230개 이상의 도구를 사용하여 디자인 소프트웨어를 조작하며, 외부의 procedural memory가 자연어 형태의 skill을 축적하고 정제합니다. 이 과정은 반복되는 subtask를 학습하고 기존 procedure를 성공 및 실패 경험에 따라 수정하며 지속적으로 발전합니다.
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