当检索到的文档竟出自AI自己之手，大模型回答会逐步退化崩塌，这份研究揭开了RAG系统被自身输出污染的风险。
arXiv:2608.22118v1 Announce Type: new Abstract: LLM responses are based on the internet (via training or RAG), and AI is now used to generate a signif…
智能体记忆系统能否跟上环境变化，这项研究给出了关键答案与实验验证。
arXiv:2608.19652v1 Announce Type: cross Abstract: As LLM-based agents are deployed for longer and higher-stakes tasks, their memory systems continue t…
用列文-柯尔莫哥洛夫复杂度度量提示词价值，给“好提示”一个可计算的理论标尺。
arXiv:2608.16438v1 Announce Type: new Abstract: In a world where valuable artifacts are increasingly created, completed, or processed by LLMs, the cen…
不只是问答，这项研究揭示了LLM在战略博弈中的推理短板，值得关注。
arXiv:2510.10813v2 Announce Type: replace Abstract: Large Language Models (LLMs) are increasingly applied to domains that require reasoning about othe…
用算子式框架管理LLM心智模型，破解长文本记忆压缩与动态更新的难题。
arXiv:2608.15451v1 Announce Type: new Abstract: Large language models process large amounts of information but usually lack an explicit mechanism for …
掩码扩散模型可任意顺序生成关卡，绕开求解器直接产出可解推箱子难题，值得一看。
arXiv:2608.15958v1 Announce Type: new Abstract: Deciding whether a Sokoban puzzle is solvable is PSPACE-complete (Culberson, 1997): solutions can be e…
经典AI研究并未消亡，而是以工具形式嵌入大模型？HN上的高质量提问，引发对AI研究范式演变的思考。
That is, classical AI has turned into tools to be used by LLMs as the need arises? So classical AI research is still continuing? Comments URL: https:/…
一篇颠覆认知的AI研究报告，用因果证据揭示AI使用对认知的即时影响，AI从业者必读
The room is unremarkable. A clean desk, a laptop, a printed sheet of arithmetic and short reading passages, and a pair of EEG electrodes resting again…
AI研究速览：聚焦评估污染与检索扩展，一窥模型评测新盲区。
Reliable provenance and graded trust Explicit reliability modeling cuts hallucination. Σ‑Mem stores symmetric competence states for peer agents, while…
揭秘AI如何写物理动画代码，新基准测试与训练方法一睹为快。
arXiv:2602.10840v2 Announce Type: replace Abstract: Large language models (LLMs) have been widely studied in areas such as mathematical reasoning, com…
直击扩散语言模型预训练与生成阶段的不匹配痛点，提出改进方法，值得算法研究者细读。
arXiv:2608.09424v1 Announce Type: new Abstract: Autoregressive language models align training and use: generation conditions on a clean prompt, and tr…
专注AI研究解读的资讯站，用前沿论文帮你理解AI如何影响认知与信念。
Article URL: https://aiunderstanding.org/news/ai-chats-emerging-conspiracy-beliefs-study Comments URL: https://news.ycombinator.com/item?id=49241569 P…
面向长时程终端任务，递归合成方法带来全新解决思路，值得关注。
arXiv:2608.05466v1 Announce Type: new Abstract: High-quality long-horizon training data for terminal agents is expensive to produce, often costing hun…
这项研究将3D建模与空间推理解耦，为AI生成更精细的3D场景提供新思路。
arXiv:2608.05242v1 Announce Type: new Abstract: In this work, we explore an alternative paradigm for spatial reasoning by explicitly disentangling 3D …
大模型类比能力多样性研究揭示AI推理新维度，或为认知智能评估提供全新视角。
arXiv:2608.03233v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated remarkable potential for analogy making, a core cogniti…
探究大模型能否跳出单一推理路径，自主发现多元异构推断，视角新颖。
arXiv:2608.02867v1 Announce Type: new Abstract: Although reinforcement learning with verifiable rewards (RLVR) has improved the performance of large l…
AI研究瓶颈转向世界模型？HN热议真实数据收集与实验指导的下一站。
It seems that generating hypotheses is no longer the main problem. The harder problem may be continuously collecting real-world data, building accurat…
让AI智能体在交互中自主进化，Living-Harness开辟智能体持续学习新范式。
arXiv:2607.26598v1 Announce Type: cross Abstract: Large language model (LLM) agents may recover from a failure within an episode or after a retry, yet…
研究MLLM生成可视化声明是源于图表数据还是模型知识，为无障碍数据理解开启新视角
arXiv:2607.25021v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) can connect visualization patterns to external causes, conseq…