analiza aiqbench
This research introduces Learning to Share (LTS), a mechanism that optimizes parallel agentic systems by selectively reusing intermediate reasoning steps across multiple teams. By training a lightweight controller with reinforcement learning to manage a global memory bank, the approach significantly reduces computational redundancy without sacrificing task performance on standard benchmarks. This work addresses critical efficiency challenges in scaling multi-agent architectures through intelligent information sharing rather than isolated execution.
Learning to Share: Selective Memory for Efficient Parallel Agentic Systems
arXiv:2602.05965v2 Announce Type: replace-cross Abstract: Agentic systems solve complex tasks by coordinating multiple agents that iteratively reason, invoke tools, and exchange intermediate results. To improve robustness and solution quality, recent approaches deploy multiple agent teams running in parallel to explore diverse reasoning trajectories. However, parallel execution comes at a significant computational cost: when different teams independently reason about similar sub-problems or execute analogous steps, they repeatedly perform substantial overlapping computation. To address these limitations, in this paper, we propose Learning to Share (LTS), a learned shared-memory mechanism for parallel agentic frameworks that enables selective cross-team information reuse while controlling context growth. LTS introduces a global memory bank accessible to all teams and a lightweight controller that decides whether intermediate agent steps should be added to memory or not. The controller is trained using stepwise reinforcement learning with usage-aware credit assignment, allowing it to identify information that is globally useful across parallel executions. Experiments on the AssistantBench and GAIA benchmarks show that LTS significantly reduces overall runtime while matching or improving task performance compared to memory-free parallel baselines, demonstrating that learned memory admission is an effective strategy for improving the efficiency of parallel agentic systems. Project page: https://joefioresi718.github.io/LTS_webpage/
Abstract:Agentic systems solve complex tasks by coordinating multiple agents that iteratively reason, invoke tools, and exchange intermediate results. To improve robustness and solution quality, recent approaches deploy multiple agent teams running in parallel to explore diverse reasoning trajectories. However, parallel execution comes at a significant computational cost: when different teams independently reason about similar sub-problems or execute analogous steps, they repeatedly perform substantial overlapping computation. To address these limitations, in this paper, we propose Learning to Share (LTS), a learned shared-memory mechanism for parallel agentic frameworks that enables selective cross-team information reuse while controlling context growth. LTS introduces a global memory bank accessible to all teams and a lightweight controller that decides whether intermediate agent steps should be added to memory or not. The controller is trained using stepwise reinforcement learning with usage-aware credit assignment, allowing it to identify information that is globally useful across parallel executions. Experiments on the AssistantBench and GAIA benchmarks show that LTS significantly reduces overall runtime while matching or improving task performance compared to memory-free parallel baselines, demonstrating that learned memory admission is an effective strategy for improving the efficiency of parallel agentic systems. Project page: this https URL
Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI)
arXiv:2602.05965 [cs.MA]
(or arXiv:2602.05965v2 [cs.MA] for this version)
https://doi.org/10.48550/arXiv.2602.05965 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 289 cuvinte
Subiecte conexe
Token Utility Is Selection-Conditioned: Coupled Selection of Prompt Context and Response Supervision for Efficient Instruction Tuning
arXiv · 22 sept. 2026
Best Policy Tracking in Reinforcement Learning
arXiv · 27 oct. 2026
PRISM: Self-Pruning Intrinsic Selection Method for Training-Free Multimodal Data Selection
arXiv · 1 sept. 2026
EvoCSFL: Surrogate-Assisted Evolutionary Client Selection for Efficient and Robust Federated Learning
arXiv · 9 iun. 2026