Enabling Memory and Reasoning in Foundation Agents

​Foundation agents, built on the backbone of LLMs, are evolving from passive responders to active thinkers—autonomously remembering, reasoning, and improving across tasks and domains.
Yet two cognitive capabilities remain crucial bottlenecks: how they remember and how they think.

In this talk, our guest Bang Liu, associate professor at the University of Montreal, will present a cognitive-inspired framework for understanding and architecting foundation agents and focus on two core pillars—memory and reasoning.

R3Mem introduces reversible memory compression to balance long-term retention with precise retrieval, enabling LLM agents to recall extended histories and interact coherently across long horizons.

​System-1.5 Reasoning breaks the dichotomy between fast heuristics and slow deliberation by creating dynamic shortcuts in latent space. It achieves CoT-level reasoning with up to 20× faster inference, bridging System-1 speed and System-2 depth.

Read the paper
​Join BuzzRobot Slack to connect with the community