一个评估AI智能体能否“重新发现”科学洞察的创新基准，ICML 2026收录，值得研究者关注。
arXiv:2602.02905v2 Announce Type: replace Abstract: Autonomous agents powered by large language models (LLMs) promise to accelerate scientific discove…
多模态大模型持续学习基准够用吗？ICML 2026 Workshop论文深度剖析现有评估体系的不足。
arXiv:2606.20961v1 Announce Type: cross Abstract: Continual adaptation is essential for multimodal large language models (MLLMs) deployed across evolv…
新协议揭穿多智能体LLM协调增益的虚假繁荣，用配对噪声底线方法精准衡量真实协同效应。
arXiv:2606.20695v1 Announce Type: cross Abstract: Multi-agent LLM coordination papers report small benchmark deltas as evidence that one architecture …
度量LLM数学定理证明能力的全新基准，填补自动形式化证明评估空白
arXiv:2606.09450v1 Announce Type: new Abstract: LLMs have recently achieved strong results on formal proving benchmarks. However, existing evaluations…
揭示深度研究代理中搜索时间污染导致基准性能虚高，为AI评估体系敲响警钟。
arXiv:2606.05241v1 Announce Type: cross Abstract: Public benchmarks enable fair and reproducible evaluation of LLM reasoning, but they become fragile …
解决LLM安全评估的两大缺陷：多数据集统一阈值与隐含操作点透明化，这套评估框架更严谨。
arXiv:2606.02959v1 Announce Type: new Abstract: Published evaluations of prompt-injection and jailbreak detectors for Large Language Models often suff…
系统评估多模态大模型在视频中跟踪实体、状态与事件变化的能力，揭示当前视频理解的关键瓶颈。
arXiv:2606.03920v1 Announce Type: new Abstract: Understanding a video requires more than recognizing isolated moments, as humans continuously track en…
压缩LLM时，准确率不是唯一指标——新基准用保形概率评估不确定性保留。
arXiv:2606.01850v1 Announce Type: new Abstract: Model compression techniques such as quantization and pruning are widely used to reduce the deployment…
新基准JuICE：专门测试LLM判断文化错误的能力，为全球部署中的文化适配提供评估标准。
arXiv:2605.26955v1 Announce Type: cross Abstract: As large language models (LLMs) are increasingly deployed to users around the world, they are integr…
首个多间隔胸片纵向推理基准，填补医学视觉问答在时序评估上的空白。
arXiv:2605.15574v1 Announce Type: new Abstract: Longitudinal chest X-ray (CXR) interpretation requires reasoning over disease evolution across multipl…