Revision aware graph method improves dynamic task reasoning efficiency
Revision-Aware Independent Agent Graphs for Dynamic Reasoning
Artificial IntelligenceComputation and Language
Summary
Sometimes tasks or questions get changed over time, and computers need to figure out what the latest correct information is before answering. The authors looked at how existing methods can waste effort by redoing too much or give wrong answers when they use outdated info. They created a new approach that treats different versions of documents separately and only checks for updates carefully, saving work while improving accuracy. They tested this on many question sets that change over time and showed their method works better and more efficiently than previous ones.
What this means in practice
- •For information retrieval engineers: Improve systems that answer evolving queries by efficiently handling document revisions for more accurate responses over time.
- •For software maintenance teams: Use revision-aware reasoning to manage and verify task outputs when source documents or requirements are frequently updated.
Authors
Yan Luo, Selim-Antoine Lali, Jeremy Moebel, Iliass Khoutaibi, Ahmadou Aidara, Mengyu Wang
Abstract
Conventional reasoning protocols present a fixed, preselected task, so they cannot test whether an agent propagates relevant updates, preserves unaffected work, or reconstructs a historical task binding. We therefore study \emph{dynamic task routing}, in which an event stream revises task bindings and a system must select the document version valid at each query time before solving it. To study this problem, we repurpose six widely used benchmarks: MMLU, MMLU-Pro, MedMCQA, MATH, GPQA, and HumanEval into 31{,}119 dynamic episodes comprising 373{,}428 temporally categorized queries. This setting exposes a central trade-off: recomputing after every event wastes work, whereas unguarded reuse returns stale conclusions. We introduce the Revision-Aware Independent Agent Graph (RIAG), a bounded multi-agent policy that separates deterministic temporal resolution from task reasoning. RIAG caches solutions by immutable document identity, starts each fresh task with two unexposed attempts, and conditionally invokes audit and repair, using at most four calls per document version. On this collection, homogeneous RIAG achieves 54.24\% joint routing-and-answer accuracy at 0.62 calls/query, compared with 32.22\% at 18.00 calls/query for the strongest comparison method; heterogeneous RIAG reaches 49.78\% at 0.63 calls/query.