# Can Agent Memory Systems Track Evolving State?

arXiv: https://arxiv.org/abs/2608.19652 · HTML: https://mlsanity.com/paper/2608.19652/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in Agent memory systems, one of 16 directions in LLM agents and tool use.
- Builds on 16 of its 30 references that are in the map.
- First posted on arXiv 20 Aug 2026.

## Placed in

- LLM agents and tool use / [Agent memory systems](https://mlsanity.com/agents/agent-memory-systems/)

## Builds on (most cited first, 16 of 16)

- [Generative Agents: Interactive Simulacra of Human Behavior](https://mlsanity.com/paper/2304.03442/) (2023)
- [τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains](https://mlsanity.com/paper/2406.12045/) (2024)
- [MemGPT: Towards LLMs as Operating Systems](https://mlsanity.com/paper/2310.08560/) (2023)
- [A-MEM: Agentic Memory for LLM Agents](https://mlsanity.com/paper/2502.12110/) (2025)
- [MultiWOZ -- A Large-Scale Multi-Domain Wizard-of-Oz Dataset for Task-Oriented Dialogue Modelling](https://mlsanity.com/paper/1810.00278/) (2018)
- [τ2-Bench: Evaluating Conversational Agents in a Dual-Control Environment](https://mlsanity.com/paper/2506.07982/) (2025)
- [Towards Scalable Multi-domain Conversational Agents: The Schema-Guided Dialogue Dataset](https://mlsanity.com/paper/1909.05855/) (2019)
- [Accelerating scientific discovery with Co-Scientist](https://mlsanity.com/paper/2502.18864/) (2025)
- [Zep: A Temporal Knowledge Graph Architecture for Agent Memory](https://mlsanity.com/paper/2501.13956/) (2025)
- [LongRoPE: Extending LLM Context Window Beyond 2 Million Tokens](https://mlsanity.com/paper/2402.13753/) (2024)
- [Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactions](https://mlsanity.com/paper/2507.05257/) (2025)
- [Measuring AI Ability to Complete Long Software Tasks](https://mlsanity.com/paper/2503.14499/) (2025)
- [TripPy: A Triple Copy Strategy for Value Independent Neural Dialog State Tracking](https://mlsanity.com/paper/2005.02877/) (2020)
- [MemoryArena: Benchmarking Agent Memory in Interdependent Multi-Session Agentic Tasks](https://mlsanity.com/paper/2602.16313/) (2026)
- [AMA-Bench: Evaluating Long-Horizon Memory for Agentic Applications](https://mlsanity.com/paper/2602.22769/) (2026)
- [STALE: Can LLM Agents Know When Their Memories Are No Longer Valid?](https://mlsanity.com/paper/2605.06527/) (2026)

## Built on by (most cited first, 7 of 7)

- [TWIST: A Proposed Benchmark for Intervention Quality in Conversational Memory, with a Human-Validated Draft-Alignment](https://mlsanity.com/paper/2609.28575/) (2026)
- [Personalized State-Transition-Aware Memory for Clinical Agents](https://mlsanity.com/paper/2609.38490/) (2026)
- [Nous: Learning and Certifying Memory Decisions Before Source Calibration](https://mlsanity.com/paper/2610.00094/) (2026)
- [Knowing the Store: What a Memory Backend Must Write Down Before an Agent Can Read It](https://mlsanity.com/paper/2610.04794/) (2026)
- [ServeLearnBench: How Well Can Agents Self-Improve from Serving Experience?](https://mlsanity.com/paper/2610.07792/) (2026)
- [Bookkeeping, Composition, or Unreachable Gold? Reading MemoryAgentBench's Conflict-Resolution Scores Against a Frozen Last-Write Resolver](https://mlsanity.com/paper/2610.09193/) (2026)
- [Suggested Is Not Superseded: A Pre-Registered Evaluation of a System-One Model as a Memory-Retirement Gate](https://mlsanity.com/paper/2610.15710/) (2026)

## Abstract (arXiv)

As LLM-based agents are deployed for longer and higher-stakes tasks, their memory systems continue to have crucial gaps. While existing memory benchmarks focus largely on recall-shaped tasks, we argue an effective memory system must track the evolving state of the world; as facts, constraints, and decisions are revised over a long interaction, answers must reflect the current state and not a superseded one. We define this capability as state tracking and instantiate it in StateMemBench, a benchmark of 234 multi-session scenarios spanning two conversation-length regimes. Its closed-pool grading scores whether an answer reflects the current state, the superseded state, or fails otherwise, separating state-tracking failures from other errors by construction. Our analysis shows that this task is challenging for existing memory systems, retrieval-augmented baselines, and long-context baselines. We then present StateMem, a state-first memory method that explicitly tracks supersession and relational dependencies, and show it improves current-state accuracy over the strongest same-backbone baseline by 1.8x (0.205 -> 0.363) on DeepSeek-V4-Flash and over the strongest memory system by 1.6x (0.149 -> 0.233) on Qwen-3.5-9B, while remaining competitive with the long-context baselines. Finally, we show the same state approach can be applied as a lightweight single-call wrapper over existing memory systems, lifting current-state accuracy by +32 to +67 points on StateMemBench across six memory and retrieval backends. A length- and cost-matched control attributes +15 to +32 of those points to state structure rather than added context.