SparseRead, AI 에이전트의 토큰 사용량 최대 92.9% 감소
Read Less, Solve More: Token-Efficient Sparse Reading of External Content for AI Agents
- 매체
- Hugging Face
- 게시
- 중요도
- 2/5
- 화제 지수
- 10
한국어 요약
웹 에이전트가 웹페이지·문서·저장소·툴 응답을 반복적으로 접근할 때, 전체 콘텐츠를 읽어 토큰과 지연 시간을 낭비하는 문제를 다뤘다. SparseRead는 에이전트 모델을 수정하지 않고 콘텐츠 노출을 제어하는 읽기 계층이다. 진화하는 요구사항 추적, 적응형 검색 백엔드, 예산 기반 게이트로 구성되며 이질적 자원을 지원한다. 6개 모델 5개 시나리오에서 총 토큰을 최대 92.9%, 실행 시간을 최대 89.0% 감소시켰고 97% 설정에서 작업 품질을 유지하거나 개선했다.
원문 발췌 · English
Web-facing AI agents repeatedly access webpages, documents, repositories, and tool responses, yet reading interfaces often expose entire objects when only sparse evidence is needed. This over-reading increases token and latency costs and can dilute relevant evidence. We present SparseRead, a reusable reading control layer that governs content admission without modifying the agent model. Its Read Intent Protocol tracks evolving needs and source coverage; Adaptive Reader Backends select how evidence is located and exposed; and the Budgeted Read Gate bounds cumulative raw exposure per episode with faithful full-view fallback. A unified Reader interface supports interchangeable evidence acquisition across heterogeneous resources. Across six models, including Claude Opus 5, and five scenarios, SparseRead reduces total tokens by up to 92.9% and wall time by up to 89.0%, preserving or improving task quality in 97% settings. On Web deep-search tasks, it reduces tokens by 31.0--43.0% and task completion time by 23.0--37.0%. Ablations validate the three designs, while integration with three agent frameworks demonstrates portability.