Providing AI agents with reliable long-term memory that does not hallucinate remains an open problem. Current approaches to memory for LLM agents -- sliding windows, summarization, embedding-based RAG, and flat fact extraction -- each reduce token cost but introduce catastrophic information loss, semantic drift, or uncontrolled hallucination about the user. The structural reason is architectural: every published memory system on the LoCoMo benchmark treats conversation as a retrieval problem over raw or lightly summarized dialogue segments, and none reports adversarial robustness, the ability to refuse questions about facts the user never disclosed. We present Synthius-Mem, a brain-inspired structured persona memory system that takes a fundamentally different approach. Instead of retrieving what was said, Synthius-Mem extracts what is known about the person: a full persona extraction pipeline decomposes conversations into six cognitive domains (biography, experiences, preferences, social circle, work, psychometrics), consolidates and deduplicates per domain, and retrieves structured facts via CategoryRAG at 21.79 ms latency. On the LoCoMo benchmark (ACL 2024, 10 conversations, 1,813 questions), Synthius-Mem achieves 94.37% accuracy, exceeding all published systems including MemMachine (91.69%, adversarial score is not reported) and human performance (87.9 F1). Core memory fact accuracy reaches 98.64%. Adversarial robustness, the hallucination resistance metric that no competing system reports, reaches 99.55%. Synthius-Mem reduces token consumption by ~5x compared to full-context replay while achieving higher accuracy. Synthius-Mem achieves state-of-the-art results on LoCoMo and is, to our knowledge, the only persona memory system that both exceeds human-level performance and reports adversarial robustness.

翻译：为AI智能体提供可靠且不产生幻觉的长时记忆仍是一个开放性问题。当前面向大语言模型智能体的记忆方案——包括滑动窗口、摘要总结、基于嵌入的RAG以及扁平化事实抽取——虽各能降低token成本，却导致灾难性信息丢失、语义漂移或对用户信息产生失控幻觉。其结构性原因在于架构层面：LoCoMo基准测试中所有已发表的记忆系统均将对话视为对原始或轻度摘要对话片段的检索问题，且无一报告对抗鲁棒性——即拒绝回答用户未披露事实相关问题的能力。本文提出Synthius-Mem，一种受脑启发的结构化人格记忆系统，采用根本性不同的方法。不同于检索"说过什么"，Synthius-Mem抽取"关于人物已知什么"：完整的人格抽取流程将对话分解为六个认知域（生平、经历、偏好、社交圈、工作、心理测量），对每个域进行整合与去重，并通过CategoryRAG以21.79毫秒延迟检索结构化事实。在LoCoMo基准测试（ACL 2024，含10组对话、1,813个问题）上，Synthius-Mem取得94.37%准确率，超越所有已发表系统——包括MemMachine（91.69%，未报告对抗得分）以及人类表现（87.9 F1）。核心记忆事实准确率达98.64%。对抗鲁棒性（竞争系统均未报告的幻觉抵抗指标）达99.55%。与全上下文回放相比，Synthius-Mem在实现更高准确率的同时将token消耗降低约5倍。Synthius-Mem在LoCoMo上达到了最先进水平，并且据我们所知，是首个既超越人类表现又报告对抗鲁棒性的人格记忆系统。