about
Learning Who Talks to Whom: Role-Aware Communication Topology Scheduling (arxiv.org)
1 point by Gedxx 28 days ago | hide | past | pdf | discuss on HN

In plain words: A controller picks which AI agents talk to whom and when, using each one's role and progress, and scores the team on the actual conversation graph. Across 9 benchmarks it beat top rivals by up to 6.19 accuracy points with over 15% fewer tokens.

Abstract · Learning Who Talks to Whom: Role-Aware Communication Topology Scheduling for Efficient LLM Multi-Agent Collaboration

Multi-agent systems driven by large language models (LLMs) rely on structured communication topologies to facilitate information exchange and collaborative reasoning. However, effectively scheduling these topologies to dynamically determine who talks to whom and when to communicate remains a fundamental challenge in balancing collective reasoning performance against exponential token overhead. Existing topology scheduling approaches either enforce static communication graphs that cannot adapt to dynamic problem-solving states, or overlook agent role complementarity while failing to explicitly incorporate the synthesized graph structure into team-level decision evaluation. To tackle these limitations, we propose ACTS, a role-aware adaptive communication topology scheduling framework for efficient LLM multi-agent collaboration. Orchestrated within a Centralized Training with Decentralized Execution (CTDE) paradigm, ACTS introduces a novel dual-attention mechanism. Specifically, a Role-conditioned Relational Attention module dynamically selects variable-cardinality interaction partners based on agent roles and evolving execution states, while a Topological Attention Mixer explicitly embeds the directed communication topology into team value decomposition for precise credit assignment. Comprehensive evaluations across 9 challenging benchmarks covering code generation, multi-step reasoning, general knowledge, and competitive mathematics demonstrate that ACTS consistently outperforms state-of-the-art baselines, boosting average accuracy by up to 6.19 percentage points while reducing token consumption by over 15%. The project website is available at https://acts-mas.github.io/ACTS/.

Haibin Li, Wenya Yu, Xuefeng Jiang, Jia Li, Xiao Han, Shuai Su, Wei Wei
arXiv:2610.25851 · cs.MA · submitted Oct 19, 2026
abstract · pdf · html · Project website: https://acts-mas.github.io/ACTS/

add comment on HN