#Agent Systems
5 articles with this tag
Active Memory: Defeating Behavioral State Decay
New research introduces an active memory agent that combats 'behavioral state decay AI' in long-horizon tasks, boosting performance by up to +8.3 pp.
Agent JIT Compilation for Web Automation
Agent just-in-time compilation revolutionizes web automation by compiling tasks into efficient code, yielding significant speed and accuracy gains.
Code as the Agent Harness
Code is evolving into the foundational 'harness' for AI agents, enabling more executable, verifiable, and stateful systems across diverse applications.
Externalizing Agent Harnesses with Language
Researchers introduce Natural-Language Agent Harnesses (NLAHs) and an Intelligent Harness Runtime (IHR) to externalize agent control logic, enabling greater transferability and scientific study.
MiniMax M2.7 Hints at AI Self-Evolution
MiniMax's M2.7 model showcases early signs of AI self-evolution, excelling in software engineering and professional tasks while driving organizational AI transformation.