首次公开多轮RAG端到端强基线：用LLM重排序+忠诚度控制，在SemEval-2026上验证有效性。
arXiv:2606.28737v1 Announce Type: new Abstract: We introduce 5ting, our system for the SemEval2026 Task 8 (MTRAGEval), which evaluates multi-turn Retr…
知识图谱仅对超出训练范围的医学知识有效提升LLM表现，揭秘结构化知识与模型能力的边界。
arXiv:2606.22419v1 Announce Type: new Abstract: A recent Nature Medicine study reports that general-purpose frontier LLMs outperform specialized retri…
两大AI评委给Agent回答高分0.85，但Agent从未打开文件，GroundEval揭穿这巨大评分偏差。
Article URL: https://tenureai.dev/writing/llm-as-judge-became-the-default-for-agent-evaluation/ Comments URL: https://news.ycombinator.com/item?id=486…
新基准WildIFEval带你从实验室走进真实世界的指令跟随能力评估
arXiv:2503.06573v3 Announce Type: replace-cross Abstract: Recent LLMs have shown remarkable success in following user instructions, yet handling instr…
手把手教你用追踪与评估调试AI代理，提升开发效率的关键技巧。
Article URL: https://medium.com/no-time/how-to-debug-ai-agents-with-traces-and-evals-a3b72e9e7c82 Comments URL: https://news.ycombinator.com/item?id=4…
一项研究揭示多智能体LLM架构不仅影响代码正确性，还会改变代码的结构复杂度。
arXiv:2606.00308v1 Announce Type: cross Abstract: Large-language-model code generation has shifted from single-shot prompting to multi-agent orchestra…
首个标准化EEG基础模型评估基准，覆盖多任务与多数据集，推动脑电AI研究规范化。
arXiv:2606.00815v1 Announce Type: new Abstract: Electroencephalography (EEG) supports a variety of brain-computer interface (BCI) tasks ranging from b…
用AI模型区分仇恨言论与LGBTQ+群体内部自我赋权言论，EVALITA挑战赛最新研究。
arXiv:2606.01298v1 Announce Type: new Abstract: The spread of hate speech has become increasingly harmful in modern digital environments, particularly…
AI评估新范式：STABLEVAL在评估中引入分歧感知与稳定性，为AI系统的可靠评价提供创新方案。
arXiv:2605.02122v2 Announce Type: replace Abstract: Human evaluation remains the primary standard for assessing modern AI systems, yet annotator disag…
首个针对OpenQASM-3的LLM训练评估数据集，突破量子电路局限，直击NISQ时代硬件需求。
arXiv:2605.30358v1 Announce Type: new Abstract: Quantum computing remains in the Noisy Intermediate-Scale Quantum (NISQ) era, where the performance is…
大模型与儿童人格碰撞，首个针对儿童个性的评估基准ChildEval问世。
arXiv:2605.27805v1 Announce Type: cross Abstract: While LLMs enable personalized chatbots, their effectiveness in child-centered personalization remai…
基于增量语义知识图谱的状态化多轮对话评估方法，提升对话系统评测的连贯性与深度。
arXiv:2605.16650v1 Announce Type: new Abstract: Evaluating multi-turn dialogue systems remains challenging because response quality depends not only o…
首次用粒子物理实验复现任务来评估AI代理，测试长程推理与科学工具使用能力，填补现有基准缺乏真实科学复杂性的空白。
arXiv:2605.13950v1 Announce Type: cross Abstract: Autonomous language-model agents are increasingly evaluated on long-horizon tool-use tasks, but exis…