揭示长上下文LLM因位置偏差导致推理失败的盲点，挑战现有基准评估体系。
arXiv:2605.23170v1 Announce Type: cross Abstract: Position-controlled evaluation is standard for retrieval tasks such as Needle-in-a-Haystack and RULE…
深入剖析AI Agent记忆结构的分类法与系统局限，为构建更智能的代理提供实证分析。
arXiv:2602.19320v2 Announce Type: replace Abstract: Agentic memory systems enable large language model (LLM) agents to maintain state across long inte…
新基准MixRea评估LLM在显式与隐式推理上的表现，揭示推理能力的短板。
arXiv:2605.20128v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly integrated into high-stakes decision-making. Inspired by…
Agentic评估 vs LLM评判：在成本、时间和质量之间如何权衡？看这篇讨论。
Comments URL: https://news.ycombinator.com/item?id=48144995 Points: 1 # Comments: 0
OpenAI最新研究揭示语言模型幻觉根源，用更优评估提升AI可靠性与安全性。
OpenAI’s new research explains why language models hallucinate. The findings show how improved evaluations can enhance AI reliability, honesty, and sa…
新方法LPDS通过保留逻辑改变实体，精准测试大模型鲁棒性，避免模型因细节变化而翻车。
arXiv:2605.15393v1 Announce Type: new Abstract: As large language models (LLMs) are increasingly deployed to perform tasks with minimal human oversigh…
现有LLM记忆评估靠最终准确率，但会掩盖关键失败模式，本文提出新视角
arXiv:2605.15384v1 Announce Type: cross Abstract: Memory plays a central role in enabling large language models (LLMs) to operate over sequential task…