Metraipolis › archive › dc_ee7aa6fd0140

CoMemBench: Benchmarking Collaborative Memory Boundaries across Multi-Agent Workflow Topologies — arXiv:2609.32192

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.32192
License (as archived): CC0-1.0
Archived: 2026-10-07T20:55:10.846Z
SHA-256 of the archived text: 5867db1e70e5cdae939c9231944d2d16b45f9dff6c2eadcb1b3acfdb2772746b
Stable URL of this page: https://metraipolis.com/documento/dc_ee7aa6fd0140

How to cite

CoMemBench: Benchmarking Collaborative Memory Boundaries across Multi-Agent Workflow Topologies — arXiv:2609.32192 — arXiv (en), https://arxiv.org/abs/2609.32192 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ee7aa6fd0140 (Metraipolis archive dc_ee7aa6fd0140, sha256 5867db1e70e5cdae939c9231944d2d16b45f9dff6c2eadcb1b3acfdb2772746b)

BibTeX

@misc{dcee7aa6fd0140,
  title = {CoMemBench: Benchmarking Collaborative Memory Boundaries across Multi-Agent Workflow Topologies — arXiv:2609.32192},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.32192},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ee7aa6fd0140; sha256 5867db1e70e5cdae939c9231944d2d16b45f9dff6c2eadcb1b3acfdb2772746b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.32192. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Multi-agent workflows require task-relevant information to be shared across agents, while irrelevant, stale, unverified, or incompatible information must remain isolated. We call this task-conditioned scope of information a collaborative memory boundary. Workflow topology determines which intermediate artifacts are applicable to which downstream workers and when they cease to be valid, thereby providing a structural stress dimension for sharing and isolation. Existing memory benchmarks primarily evaluate retention and retrieval, whereas multi-agent benchmarks emphasize coordination and end-to-end completion, leaving topology-conditioned memory boundaries largely unmeasured. We introduce CoMemBench, an execution-grounded benchmark for collaborative memory sharing and isolation across multi-agent workflow topologies. It constructs 800 composite workflows across four domains from source-grounded dependency graphs, with node-local specifications, verifiable artifact handoffs, native evaluators, and matched isolation challenges. CoMemBench measures workflow completion, verified node progress, required-handoff reliability, isolation robustness, and token cost. Experiments reveal a sharing-isolation trade-off: broader context improves information availability but can weaken isolation, while system rankings shift across topologies and artifact violations.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.32192 — 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 2026-10-07T20:29:00.723Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).