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.
Read the full articleYou might also wanna read
Proactive Memory Agent Improves Long-Horizon AI Task Performance by Addressing Behavioral State Decay
In long-horizon tasks, decision-relevant state is often scattered across an expanding trajectory, while the action agent must surface it and
arxiv.org·6d agoProactive Memory Agent Improves Long-Horizon AI Task Performance by Addressing Behavioral State Decay
In long-horizon tasks, decision-relevant state is often scattered across an expanding trajectory, while the action agent must surface it and
arxiv.org·6d ago
Oracle Agent Memory: A Database-Native Memory Substrate for Long-Horizon AI Agents
Agent memory is a systems problem for long-horizon agents. Practical deployments require retention of task state across extended conversatio
arxiv.org·5h ago
AI Memory: The Key to Personalized Agents
AI agents need more than extended prompts for effective memory. An architecture for structured, selective, and adaptive memory could transfo
Technical Challenges and Solutions for Long-Running AI Agents
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
anthropic.com·7mo ago
Agent Memory Is Distributed State Management, Not Magic
Agent memory is not magic. It is distributed state with caches, logs, consistency windows, synchronization, and memory curation.
AI memory systems can degrade model performance and increase sycophancy, new research finds
New research suggests that AI memory systems can degrade model performance and encourage sycophantic tendencies.
techcrunch.com·1mo ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.