How asynchrony in search agents changes evidence gathering choices
The Scheduler Is a Hidden Retriever: How Asynchrony Shapes Evidence Acquisition in Search Agents
Artificial IntelligenceComputation and LanguageInformation Retrieval
Summary
When computer programs search for answers, the timing of when they get clues can change the decisions they make, even if the clues themselves are the same. The authors found that waiting for all information before deciding isn't always best; early timing can cause lasting differences in what information agents seek next. They created a new approach called Confluent Evidence Acquisition, letting agents reason locally as evidence arrives and reducing disagreement caused by timing. This method helps agents make decisions more consistently and quickly without losing accuracy.
What this means in practice
- •For ai system developers: Implement controlled timing mechanisms in language-model agents to reduce inconsistencies caused by asynchronous evidence arrival.
- •For software engineers building search tools: Use Confluent Evidence Acquisition to improve the speed and consistency of asynchronous query processing in complex search systems.
Authors
Ruiyang Ren, Yuhao Wang, Yinan Zhang, Chunyan Miao
Abstract
Language-model agents increasingly solve complex tasks through parallel reasoning and asynchronous information gathering. Controlled replay across two model families and three benchmarks shows that delivery timing redirects search decisions made before all results arrive, despite identical initial evidence. Acquisition disagreement reaches up to 24 percentage points above same-order repeat variation. This dependence persists after all initial evidence arrives and through a second search, turning temporary differences in what agents have seen into lasting differences in what they acquire. An acquisition-regret bound with a matching construction shows that invariance at a fixed early arrival count can limit the use of evidence already received. We introduce Confluent Evidence Acquisition (CEA), which instead releases local reasoning when each subquestion's evidence is complete and allocates shared search independently of completion order. For a fixed plan and stable conditional response law, we prove joint invariance of decisions, acquisitions, and answers. With question-only plans across three benchmarks, CEA brings acquisition disagreement to within 1.0 percentage point of repeat variation and achieves accuracy comparable to complete-input controls, improving over early commitment by up to 12.6 percentage points. Replacing CEA's fixed allocation with first-ready allocation under contention adds 21.6 percentage points of acquisition disagreement, while local query-target disagreement remains near repeat variation. Timed comparisons show that CEA reaches decisions earlier and responds faster than the same program waiting for all initial evidence.