Metraipolis › archive › dc_c245dcd4ec5b

Bookkeeping, Composition, or Unreachable Gold? Reading MemoryAgentBench's Conflict-Resolution Scores Against a Frozen Last-Write Resolver — arXiv:2610.09193

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.09193
License (as archived)
CC0-1.0
Archived
2026-10-08T11:09:04.742Z
SHA-256 of the archived text
d0de8b1b7e6f9fec232fe1cf17d8b7bfffc9d80fb388e83949eee0211cefae9c
Stable URL of this page
https://metraipolis.com/documento/dc_c245dcd4ec5b

How to cite

Bookkeeping, Composition, or Unreachable Gold? Reading MemoryAgentBench's Conflict-Resolution Scores Against a Frozen Last-Write Resolver — arXiv:2610.09193 — arXiv (en), https://arxiv.org/abs/2610.09193 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c245dcd4ec5b (Metraipolis archive dc_c245dcd4ec5b, sha256 d0de8b1b7e6f9fec232fe1cf17d8b7bfffc9d80fb388e83949eee0211cefae9c)
BibTeX
@misc{dcc245dcd4ec5b,
  title = {Bookkeeping, Composition, or Unreachable Gold? Reading MemoryAgentBench's Conflict-Resolution Scores Against a Frozen Last-Write Resolver — arXiv:2610.09193},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.09193},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c245dcd4ec5b; sha256 d0de8b1b7e6f9fec232fe1cf17d8b7bfffc9d80fb388e83949eee0211cefae9c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09193. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

MemoryAgentBench's Conflict Resolution split is read as measuring "selective forgetting". We execute the benchmark's own rule - the newest statement about a fact wins - as a zero-learning resolver frozen on one of the four fact lists. Under the official metric the rule answers 80.25% of the questions (74.5% on the three held-out lists). Of the rest, 67 items have a released gold that the last-write graph cannot reach but overwritten statements would ("The capital of India is New Delhi." superseded by "The capital of India is Grosseto."; gold New Delhi); such items are a third of the multi-hop questions at 262K. Two long-context models and our pre-registered approximate re-implementation of the benchmark's BM25 agent, one retained run per item and outcomes only, score 84.7%, 82.6% and 41.6% on the items the rule solves against 10.4%, 11.9% and 6.0% on those 67. The failures are a reachability split plus a small parser-scope residual; the per-item split, not the aggregate, is the unit at which a score here can be read.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.09193 — 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 2026-10-08T10:14:05.915Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).