analiza aiqbench
This research introduces CoWeaver, an algorithm designed to facilitate collaboration between human scientists and LLM-based agents by addressing the need for bidirectional matching and decision interpretability in scientific workflows. The system combines exploration strategies with greedy selection mechanisms to effectively fill capability gaps while maintaining uncertainty-aware estimates of agent performance over time. By outperforming baseline methods on multiple metrics, this work advances the practical integration of autonomous AI into complex human-agent networks rather than just standalone tasks.
CoWeaver: A Bi-directional, Learnable and Explainable Matching Engine for Mixed Human-Agent Science Collaboration
arXiv:2607.15545v1 Announce Type: cross Abstract: LLM-based agents excel at writing articles, coding and information retrieval. However, they fail to form strong collaborations within the scientific community due to the bidirectional, dynamic nature of the problem and a high demand of decision interpretability. We proposed COWEAVER, a bidirectional, learnable and explainable algorithm to match scientists and form strong collaborations within a human-agent network. COWEAVER matches candidates and requesters through filling capability gaps and filters candidates through a two-stage ranking step. Finally, the model explores newcomers by maintaining uncertainty-aware capability estimates and updating them through requester's feedback. We show that the selection mechanism of combining both exploration (UCB) and greedy of COWEAVER exceeds the greedy-only mechanism - the analytical best solution - on 6 out of the 20 tasks and performed on par with the greedy-only mechanism in terms of selecting the best candidate. We compared COWEAVER baselines in terms of matching quality and efficiency. COWEAVER outperforms baselines on all metrics.
Abstract:LLM-based agents excel at writing articles, coding and information retrieval. However, they fail to form strong collaborations within the scientific community due to the bidirectional, dynamic nature of the problem and a high demand of decision interpretability. We proposed COWEAVER, a bidirectional, learnable and explainable algorithm to match scientists and form strong collaborations within a human-agent network. COWEAVER matches candidates and requesters through filling capability gaps and filters candidates through a two-stage ranking step. Finally, the model explores newcomers by maintaining uncertainty-aware capability estimates and updating them through requester's feedback. We show that the selection mechanism of combining both exploration (UCB) and greedy of COWEAVER exceeds the greedy-only mechanism - the analytical best solution - on 6 out of the 20 tasks and performed on par with the greedy-only mechanism in terms of selecting the best candidate. We compared COWEAVER baselines in terms of matching quality and efficiency. COWEAVER outperforms baselines on all metrics.
Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
arXiv:2607.15545 [cs.MA]
(or arXiv:2607.15545v1 [cs.MA] for this version)
https://doi.org/10.48550/arXiv.2607.15545 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
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 · 255 cuvinte
Subiecte conexe
PFedMA: Who Should Join Next as Federated Multi-Agent Orchestration Evolves?
arXiv · 15 oct. 2026
DiffUNet^2: Bidirectional Prediction, Probabilistic Generation and Collaborative Visual Discovery for Scientific Data
arXiv · 3 iun. 2026
Long-Term Behavioral Evaluation for Trusted Collaborator Selection via Bidirectional Mamba
arXiv · 27 aug. 2026
Augmented Reality Interfaces for Human-Robot Collaboration: Development of a ROS 2-Based Sensor Streaming Framework and Validation via SLAM Algorithms
arXiv · 28 sept. 2026