MASCOT: Towards Multi-Agent Socio-Collaborative Companion Systems

2026-01-20

AI Summary

This work proposes the first general-purpose multi-agent companionship framework that jointly preserves role fidelity and enables effective group collaboration, addressing the prevalent issues of role collapse and social sycophancy that lead to homogeneous and unconstructive dialogues. The framework employs a two-tier optimization mechanism: at the lower level, role-aware behavior alignment driven by Reinforcement Learning from AI Feedback (RLAIF) ensures individual role consistency; at the upper level, a meta-strategy-guided collaborative optimization promotes diverse and efficient interactions through a social contribution reward. Evaluated in psychological support and workplace scenarios, the proposed approach significantly outperforms existing systems, achieving a 14.1-point improvement in role consistency and a 10.6-point gain in social contribution.

Technology Category

• Multiagent Systems: Coordination and Collaboration
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Intelligent Robots: Multi-Robot Systems

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Multi-agent systems (MAS) have recently emerged as promising socio-collaborative companions for emotional and cognitive support. However, these systems frequently suffer from persona collapse--where agents revert to generic, homogenized assistant behaviors--and social sycophancy, which produces redundant, non-constructive dialogue. We propose MASCOT, a generalizable framework for multi-perspective socio-collaborative companions. MASCOT introduces a novel bi-level optimization strategy to harmonize individual and collective behaviors: 1) Persona-Aware Behavioral Alignment, an RLAIF-driven pipeline that finetunes individual agents for strict persona fidelity to prevent identity loss; and 2) Collaborative Dialogue Optimization, a meta-policy guided by group-level rewards to ensure diverse and productive discourse. Extensive evaluations across psychological support and workplace domains demonstrate that MASCOT significantly outperforms state-of-the-art baselines, achieving improvements of up to +14.1 in Persona Consistency and +10.6 in Social Contribution. Our framework provides a practical roadmap for engineering the next generation of socially intelligent multi-agent systems.

Problem

• persona collapse
• social sycophancy
• multi-agent systems
• socio-collaborative companions
• dialogue redundancy

Innovation

• Multi-agent systems
• Persona-Aware Behavioral Alignment
• Collaborative Dialogue Optimization
• RLAIF
• Socio-collaborative companions