Chain-of-thought (CoT) reasoning allows humans to inspect how large language models reach their answers, and oversee mod…
为什么值得关注：属于技术进展，可能扩散到其他领域
Recent advances in autoregressive video generation have improved temporal consistency over extended durations, yet inter…
为什么值得关注：属于技术进展，可能扩散到其他领域
A reasoning model asked whether a causal effect is recoverable from observational data can fail in two ways: it refuses…
为什么值得关注：属于技术进展，可能扩散到其他领域
General-purpose Text-to-SQL systems achieve strong performance on academic benchmarks like Spider and BIRD, where schema…
为什么值得关注：内容与关注主题相关
Explainers for Graph Neural Networks (GNNs) are commonly evaluated by their plausibility, i.e., how well their explanati…
为什么值得关注：属于政府/政策信息，对教育或研究方向可能产生影响
Generating progressively harder reasoning problems requires synthesis procedures that adapt as the task distribution evo…
为什么值得关注：属于技术进展，可能扩散到其他领域
Cephalonauts One is a whole-brain 3 Tesla (3T) functional magnetic resonance imaging (fMRI) dataset recorded while subje…
为什么值得关注：内容与关注主题相关
Financial AI agents must do more than retrieve facts: investment workflows require correct quantitative execution, relia…
为什么值得关注：属于技术进展，可能扩散到其他领域
Contactless heart-rate sensing with millimeter-wave (mmWave) radar requires assessing whether individual measurements su…
为什么值得关注：属于政府/政策信息，对教育或研究方向可能产生影响
We introduce HyperBrowseComp, a multilingual and multimodal browsing benchmark comprising 423 manually authored and huma…
为什么值得关注：属于技术进展，可能扩散到其他领域
Diffusion Transformers (DiTs) can generate high-quality images and videos, but generating each sample requires multiple…
为什么值得关注：属于技术进展，可能扩散到其他领域
Security benchmarks for LLM-based agents often report the attack success rate (ASR) as a measure of model robustness and…
为什么值得关注：属于技术进展，可能扩散到其他领域
Understanding and assessing natural hazards is essential for disaster preparedness and risk reduction.Recent advances in…
为什么值得关注：属于技术进展，可能扩散到其他领域
Reinforcement learning (RL) is increasingly used for financial optimal-control problems when complex dynamics make analy…
为什么值得关注：属于政府/政策信息，对教育或研究方向可能产生影响
Continuous emotion regression estimates moment-to-moment valence and arousal while a viewer watches a video.In familiar-…
为什么值得关注：提供了新的研究结果
The recent wave of one-step generative models, which compress the multi-step trajectory of diffusion via either distilla…
为什么值得关注：提供了新的研究结果
Designing a scientific instrument tests whether language-model agents can do physics rather than recall it, provided the…
为什么值得关注：属于技术进展，可能扩散到其他领域
Terminal-using agents benefit from reinforcement learning (RL) in coding, debugging, and other multi-step terminal tasks…
为什么值得关注：属于政府/政策信息，对教育或研究方向可能产生影响
Camera-controlled video models are rapidly advancing toward long generation horizons and complex camera control.A key fa…
为什么值得关注：属于技术进展，可能扩散到其他领域
Large language models (LLMs) are increasingly used to support legal practice, education, and research, yet their reliabi…
为什么值得关注：提供了新的研究结果