When Does Selection Replace Extraction? A Pre-Registered Test of Agent Memory with a Typed Decision Model

• 类型：arxiv
• 标识：2609.34227
• 链接：https://arxiv.org/abs/2609.34227
• 主分类：agent
• 形态：method
• TLDR：Does conversational memory need LLM-extracted facts, or is selecting the right raw turns enough? Published results disagree. Extraction-based systems report gains from distilled facts. Recent studies find raw history with good ranking does as well, but disagree about whether ranking matters. We ran a pre-registered study on held-out LoCoMo conversations and LongMemEval. At a tight budget on LoCoMo, raw turns selected by a single call to Jev, a typed decision model, are non-inferior to an LLM-extraction memory (one-sided 95% bound -3.0 points against a -5-point margin). Blind human grading narr
• 待LLM分类：否
• 标题中文：选择何时取代提取？基于类型化决策模型的 Agent 记忆预注册测试
• TLDR中文：[TLDR 中文] 对话记忆是否需要 LLM 抽取的事实，还是仅需选取合适的原始轮次？已发表结果存在分歧：基于抽取的方法报告了来自精炼事实的增益，而近期研究表明带有良好排序的原始历史亦可达到相当效果，但在排序是否关键这一问题上仍有分歧。我们在预注册研究中，对留出的 LoCoMo 对话与 LongMemEval 进行了评估。在 LoCoMo 的紧预算下，通过单次调用 Jev（一种类型化决策模型）选取的原始轮次，相对 LLM 抽取式记忆呈非劣效（针对 -5 百分点边界的单侧 95% 边界为 -3.0 百分点）。盲法人工评分...
• 来源文件：
• /inbox/tom/_candidates/2026-10-07-agent-rag-longcontext-candidates.json