Long Context Modeling
PulseAugur coverage of Long Context Modeling — every cluster mentioning Long Context Modeling across labs, papers, and developer communities, ranked by signal.
1 天有情绪数据
-
研究表明训练数据课程能微调强化学习代理的专业化
arXiv上的一项新研究探讨了不同的训练数据课程如何影响旨在与大型语言模型（LLM）和外部记忆库协同工作的强化学习（RL）代理的性能。研究发现，训练数据的构成显著影响代理的专业化，而非普遍提升性能。结合不同基准的混合课程产生了最佳的总体结果，而仅在狭窄的域外数据集上训练则特别提高了时间推理能力。
-
DeferMem框架通过强化学习增强LLM长期记忆问答能力
研究人员开发了DeferMem，一个旨在改进大型语言模型在处理长期对话记忆时的问答能力的新框架。该系统将过程分为初步的广泛候选检索和随后的条件查询证据蒸馏阶段。DeferMem利用一种名为DistillPO的强化学习算法，将检索到的信息提炼成简洁、相关的证据，在准确性和效率方面优于现有方法。
-
HyMem architecture boosts LLM agent memory efficiency by 92.6%
Researchers have developed HyMem, a novel hybrid memory architecture designed to improve the efficiency and effectiveness of large language model (LLM) agents in long-context scenarios. HyMem utilizes a dual-granular st…
-
EviMem improves conversational memory retrieval with evidence gap diagnosis
Researchers have developed EviMem, a novel framework for improving long-term conversational memory by iteratively refining retrieval queries. Unlike previous methods, EviMem explicitly identifies and addresses "evidence…
-
AdaMem: Adaptive User-Centric Memory for Long-Horizon Dialogue Agents
Multiple research papers released on arXiv propose novel frameworks for enhancing the memory capabilities of Large Language Model (LLM) agents. These approaches aim to overcome limitations in handling long-term conversa…
-
ZenBrain AI integrates 15 neuroscience models for advanced memory architecture
Researchers have introduced ZenBrain, a novel 7-layer memory architecture for AI systems inspired by neuroscience principles. Unlike existing AI memory systems that use engineering metaphors, ZenBrain integrates concept…
-
AI memory systems mimic human forgetting with 52% retention
Researchers have developed a new AI memory system that mimics the human brain's forgetting process, achieving 52% memory persistence. This system, named YourMemory and LoCoMo, is slated for release in 2026. The developm…
-
新的人工智能智能体记忆系统利用视觉和语义方法处理长时程任务
两篇新的研究论文提出了用于自主人工智能智能体的创新记忆架构，以处理长时程任务。OCR-Memory 利用智能体经验的视觉表示，以最小的开销存储广泛的历史记录，并通过定位和转录方法检索信息以减少幻觉。Memanto 引入了一个具有信息论检索引擎的类型化语义记忆系统，通过消除摄入延迟并与基于图的系统相比降低操作复杂性，在基准测试中实现了最先进的准确性。