Metraipolis › archive › dc_7ceda346828d

Can Agent Memory Systems Track Evolving State? — arXiv:2608.19652

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.19652
License (as archived)
CC0-1.0
Archived
2026-10-18T11:39:28.931Z
SHA-256 of the archived text
538dcfa1a86ef60e987419c3bbf6e8a6bd9832267302c050591ed854b74e8afa
Stable URL of this page
https://metraipolis.com/documento/dc_7ceda346828d

How to cite

Can Agent Memory Systems Track Evolving State? — arXiv:2608.19652 — arXiv (en), https://arxiv.org/abs/2608.19652 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7ceda346828d (Metraipolis archive dc_7ceda346828d, sha256 538dcfa1a86ef60e987419c3bbf6e8a6bd9832267302c050591ed854b74e8afa)
BibTeX
@misc{dc7ceda346828d,
  title = {Can Agent Memory Systems Track Evolving State? — arXiv:2608.19652},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.19652},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7ceda346828d; sha256 538dcfa1a86ef60e987419c3bbf6e8a6bd9832267302c050591ed854b74e8afa},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.19652. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

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facts, constraints, and decisions are revised over a long interaction, answers must reflect the current state and not a superseded one.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 Its closed-pool grading scores whether an answer reflects the current state, the superseded state, or fails otherwise, separating state-tracking failures from other errors by construction.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.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-21 in cs.CL).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.19652 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T11:38:27.945Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).