Tag
The article argues that autonomous AI systems should prioritize justified trust as a first-layer objective, emphasizing transparent and ethical decision-making to maintain cooperative environments in human society.
This paper explores the trade-off between memory and peer communication in bounded agents, proposing a 'remembering–signaling frontier' to optimize information allocation for task performance under resource limits.
Google DeepMind, together with Schmidt Sciences, the Cooperative AI Foundation, ARIA, and Google.org, announces up to $10M in funding for multi-agent AI safety research to study emergent behaviors and risks of interacting AI agents.
SMAC-Talk is a new benchmark that extends the StarCraft Multi-Agent Challenge to evaluate LLM-based agents in cooperative multi-agent environments with natural language communication. It includes scenarios with deceptive communicators and benchmarks agents using models from the Qwen3.5 family to study how reasoning, memory, and scale affect coordination.
The paper introduces Diamond Attention, a method for multi-agent reinforcement learning that uses structured randomness to break symmetry and enable role differentiation among homogeneous agents, achieving perfect coordination in symmetric tasks like the XOR game.