{"slug": "apex-distillation-of-agent-procedural-experience-for-adaptive-deep-research", "title": "APEx: Distillation of Agent Procedural Experience for Adaptive Deep Research Question Answering", "summary": "Researchers propose APEx, a hierarchical experience utilization framework that organizes interaction history into instance-level trajectory memories and category-level procedural skills, coupled through a closed-loop Executor-Distiller-Planner architecture. On 7 benchmarks, 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, according to the arXiv paper (2609.02253v1).", "body_md": "arXiv:2609.02253v1 Announce Type: new\nAbstract: 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.", "url": "https://wpnews.pro/news/apex-distillation-of-agent-procedural-experience-for-adaptive-deep-research", "canonical_source": "https://www.machinebrief.com/news/apex-distillation-of-agent-procedural-experience-for-adaptiv-30tj", "published_at": "2026-09-03 04:00:00+00:00", "updated_at": "2026-09-03 06:22:21.517975+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-agents", "ai-research"], "entities": ["APEx", "GPT-5.4", "arXiv"], "alternates": {"html": "https://wpnews.pro/news/apex-distillation-of-agent-procedural-experience-for-adaptive-deep-research", "markdown": "https://wpnews.pro/news/apex-distillation-of-agent-procedural-experience-for-adaptive-deep-research.md", "text": "https://wpnews.pro/news/apex-distillation-of-agent-procedural-experience-for-adaptive-deep-research.txt", "jsonld": "https://wpnews.pro/news/apex-distillation-of-agent-procedural-experience-for-adaptive-deep-research.jsonld"}}