Multi-agent LLM systems spontaneously develop power law distributions in knowledge and influence, mirroring human intellectual hierarchies. Agent societies exhibit emergent specialization and social stratification. First empirical evidence of collective social dynamics beyond individual agent capabilities.
SkillClaw enables LLM agent skills to continuously evolve through collective cross-user interaction experiences via an autonomous 'agentic evolver' that refines and updates skills, achieving +42.1% improvement. Treats agent capabilities as living artifacts that improve through collective use rather than static functions, representing a shift toward learning agent ecosystems.
WORC (Weak-link Optimization for Reasoning and Collaboration) improves multi-agent LLM frameworks by systematically identifying and reinforcing performance-limiting agents rather than only enhancing high-capability agents. Addresses reasoning instability where individual agent errors amplify through collaboration, grounded in the weak-link principle.
Proposes "agentic microphysics" methodology for analyzing safety risks that emerge from structured interactions between AI agents rather than individual model behavior. The framework bridges the gap between single-agent analysis and aggregate outcomes by focusing on communication, observation, and mutual influence mechanisms that drive population-level risks.
Comprehensive survey organizing agentic reasoning along three dimensions: foundational (planning, tool use, search), self-evolving (feedback, memory, adaptation), and collective multi-agent reasoning. Distinguishes in-context reasoning from post-training reasoning and provides unified taxonomy bridging thought and action across science, robotics, healthcare, and mathematics.