Metraipolis › archive › dc_eafcd2744c77
APEx: Distillation of Agent Procedural Experience for Adaptive Deep Research Question Answering — arXiv:2609.02253
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.02253
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T03:26:20.000Z
- SHA-256 of the archived text
c38ed87857b9b56df218152f13783a72ef96908be2e34fac2ee1b6427a5bdd91- Stable URL of this page
https://metraipolis.com/documento/dc_eafcd2744c77
How to cite
APEx: Distillation of Agent Procedural Experience for Adaptive Deep Research Question Answering — arXiv:2609.02253 — arXiv (en), https://arxiv.org/abs/2609.02253 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_eafcd2744c77 (Metraipolis archive dc_eafcd2744c77, sha256 c38ed87857b9b56df218152f13783a72ef96908be2e34fac2ee1b6427a5bdd91)
BibTeX
@misc{dceafcd2744c77,
title = {APEx: Distillation of Agent Procedural Experience for Adaptive Deep Research Question Answering — arXiv:2609.02253},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.02253},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_eafcd2744c77; sha256 c38ed87857b9b56df218152f13783a72ef96908be2e34fac2ee1b6427a5bdd91},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02253. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Deep research agents augment large language models with external tools to answer complex, long-horizon questions through multi-turn reasoning. Learning from prior experience is crucial for continual improvement, yet existing methods either retrieve verbose task-specific traces that burden decision-making, or distill procedural skills that remain decoupled from downstream policy adaptation. We propose APEx, a hierarchical experience utilization framework that organizes interaction history into instance-level trajectory memories and category-level procedural skills, and couples them through a closed-loop architecture of Executor, Distiller, and Planner. The three modules are optimized via a three-stage alternating GRPO training paradigm, enabling reward-guided skill distillation rather than fixed-prompt generation. At test time, distilled skills serve as procedural priors for online Planner adaptation through skill-guided test-time reinforcement learning, allowing ground-truth-free self-improvement with skill-alignment regularization to prevent policy drift. Experiments on 7 benchmarks demonstrate that APEx achieves state-of-the-art performance, surpassing GPT-5.4 by 14.7 points and the strongest memory-augmented baseline by 3.0 points. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-03 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2609.02253 — 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 2026-10-17T03:13:12.436Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).