Metraipolis › archive › dc_e24abb7d4a4d
CONTRAMEM: Learning Self-Evolving Procedural Memory from Contrasting Multi-Model Trajectories — arXiv:2608.22533
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.22533
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T07:41:29.279Z
- SHA-256 of the archived text
d72b298e84ffb7c7c93f4f35b1ba043a4f96ef739d4748cfdbf6331e32cc86d7- Stable URL of this page
https://metraipolis.com/documento/dc_e24abb7d4a4d
How to cite
CONTRAMEM: Learning Self-Evolving Procedural Memory from Contrasting Multi-Model Trajectories — arXiv:2608.22533 — arXiv (en), https://arxiv.org/abs/2608.22533 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e24abb7d4a4d (Metraipolis archive dc_e24abb7d4a4d, sha256 d72b298e84ffb7c7c93f4f35b1ba043a4f96ef739d4748cfdbf6331e32cc86d7)
BibTeX
@misc{dce24abb7d4a4d,
title = {CONTRAMEM: Learning Self-Evolving Procedural Memory from Contrasting Multi-Model Trajectories — arXiv:2608.22533},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.22533},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e24abb7d4a4d; sha256 d72b298e84ffb7c7c93f4f35b1ba043a4f96ef739d4748cfdbf6331e32cc86d7},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.22533. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Autonomous computer-use agents are increasingly applied to long-horizon tasks requiring coordinated application calls, persistent state tracking, and verifier-sensitive writes, yet they remain prone to procedural failures: misreading application state, tool semantics, or task progress. Procedural memory promises more consistent decisions and less redundant exploration, but constructing high-quality memory without model training remains challenging. We introduce CONTRAMEM, a source-flexible, training-free framework for self-evolving procedural memory that treats same-task outcome variation as supervision: differences in correctness, efficiency, recovery, and failure modes expose outcome-relevant procedural distinctions, distilled into a compact bank of app-level Function Cards and task-level Skill Cards that evolves through localized curation rather than append-only accumulation or whole-bank rewriting. On held-out GAIA2/ARE computer-use tasks, CONTRAMEM more than doubles the success rate across the three source-model targets (26.2% to 55.3%), with consistent per-model gains (GPT-5.5: 27.5 to 61.0; Claude Sonnet 4.6: 28.0 to 52.5; DeepSeek V4 Pro: 23.0 to 52.5). The same bank transfers unchanged to the unseen Qwen3.7 Plus (18.5 to 35.5), indicating transferable procedural knowledge rather than model-specific behavior. The same construction carries over unchanged to AppWorld, beating both no memory and its own single-source self-memory variant for all three mid-tier agents on both public test splits. Under a matched trajectory budget, heterogeneous multi-model trajectories yield stronger memory than self- or same-model multi-rollout memory: the margin comes from contrastive behavioral diversity, not stronger source agents or more sampling. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-25 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.22533 — 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 2026-10-18T07:37:52.725Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).