Critic-free reinforcement fine-tuning (RFT) for agentic large language models is often done through GRPO-style methods,…
为什么值得关注：属于技术进展，可能扩散到其他领域
Neural surrogate models for Partial Differential Equations (PDEs) on unstructured 3D geometries are often limited by poo…
为什么值得关注：属于技术进展，可能扩散到其他领域
We understand little about how capabilities acquired in one language carry over to another, or what governs this transfe…
为什么值得关注：属于技术进展，可能扩散到其他领域
Attack Path (AP) modeling is fundamental to cybersecurity analysis, where the Planning Domain Definition Language (PDDL)…
为什么值得关注：为行业报告，可提供结构化数据
Typed decision models return choices or distributions over answer options supplied at request time.Accuracy with complet…
为什么值得关注：属于技术进展，可能扩散到其他领域
Few-step neural text-to-speech models often rely on short- ened diffusion or flow-matching schedules, or on distillation…
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
Forest point cloud segmentation is fundamental for fine-grained 3D forest scene understanding, yet remains challenging d…
为什么值得关注：属于技术进展，可能扩散到其他领域
Epistemic uncertainty should decrease as additional information about the data-generating process (DGP) becomes availabl…
为什么值得关注：属于技术进展，可能扩散到其他领域
Suspicion is an important, yet elusive concept in anti-money laundering and counter-terrorist financing (AML/CFT), which…
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
Efficient request scheduling is increasingly important for reducing completion time in large language model (LLM) servin…
为什么值得关注：属于技术进展，可能扩散到其他领域
Online data selection has demonstrated substantial efficiency gains for LLM pretraining by training on the most valuable…
为什么值得关注：涉及教育实践，可对照自己的研究场景
A targeted adversarial perturbation can drive a vision-language model's (VLM's) teacher-forced training loss for a fixed…
为什么值得关注：涉及教育实践，可对照自己的研究场景
Recent progress in multimodal, high-dimensional learning has enabled foundation models to process heterogeneous, large-s…
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
Post-training with verifiable rewards can induce reward hacking, motivating the use of monitors within the training obje…
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
Accurate colon segmentation from CT images is essential for colorectal disease analysis, yet deep learning based methods…
为什么值得关注：内容与关注主题相关
Many reinforcement-learning (RL) problems are non-stationary yet structured and can be decomposed into phases, each with…
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
Long-horizon mobile manipulation presents significant challenges due to compounding execution errors and capacity interf…
为什么值得关注：属于技术进展，可能扩散到其他领域
We introduce AREX, a training-free sampler for pretrained flow matching models that uses the target mean and covariance…
为什么值得关注：属于技术进展，可能扩散到其他领域
Adversarial patches can disrupt Vision-Language-Action (VLA) models by manipulating visual observations, leading to fail…
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
Mechanistic edits (ablations, weight edits, activation steering) are the standard tools for unlearning a harmful capabil…
为什么值得关注：属于技术进展，可能扩散到其他领域