Metraipolis › archive › dc_90f0ddcd0a6b
Stashbird: Efficient Speaker-Indexed Memory for Conversational Agents — arXiv:2609.34242
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.34242
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T20:57:34.213Z
- SHA-256 of the archived text
f2fc9385672bc6bf48a1d18cda4058cb9166ef729a2f6896bc11d177a11a01e7- Stable URL of this page
https://metraipolis.com/documento/dc_90f0ddcd0a6b
How to cite
Stashbird: Efficient Speaker-Indexed Memory for Conversational Agents — arXiv:2609.34242 — arXiv (en), https://arxiv.org/abs/2609.34242 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_90f0ddcd0a6b (Metraipolis archive dc_90f0ddcd0a6b, sha256 f2fc9385672bc6bf48a1d18cda4058cb9166ef729a2f6896bc11d177a11a01e7)
BibTeX
@misc{dc90f0ddcd0a6b,
title = {Stashbird: Efficient Speaker-Indexed Memory for Conversational Agents — arXiv:2609.34242},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.34242},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_90f0ddcd0a6b; sha256 f2fc9385672bc6bf48a1d18cda4058cb9166ef729a2f6896bc11d177a11a01e7},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34242. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
AI agents require memory that preserves information across user-agent exchanges, user-to-user conversations, and group conversations with or without agent participation, while supporting updates as evidence changes or is removed. We present Stashbird, an agent memory system that links source episodes to derived memory state through explicit provenance. Stashbird organizes memory into episodic records, semantic relations, community summaries, and persisted graph state, with lifecycle operations for incremental updates and episode-level deletion. We evaluate question-answering accuracy and model-facing workload across four long-term memory benchmarks. On LoCoMo, Stashbird uses 76.4x fewer ingestion prompt tokens than Graphiti. Compared with reproduced Hindsight on the same benchmark, it uses 8.1x fewer retrieval prompt tokens, with accuracy 1.6 percentage points lower. It achieves higher accuracy than Hindsight on LongMemEval-S and GroupMemBench and comparable accuracy on EverMemBench. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.34242 — 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 2026-10-07T20:29:01.487Z 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).