CMuon: Accelerating and Stabilizing Diffusion Transformer Training via Chunked Momentum Orthogonalization
Diffusion Transformers (DiTs) have achieved state-of-the-art (SOTA) performance in visual generative modeling, yet th...
每天自动聚合 AI 领域最新动态
Diffusion Transformers (DiTs) have achieved state-of-the-art (SOTA) performance in visual generative modeling, yet th...
Can scientific abduction occur without continuous sensorimotor embodiment? Recent arguments in AI and philosophy of s...
Sequential decision-making in real-world applications often involves uncertainty about the environment's model. Uncer...
The most capable AI deployments are not single models but ensembles of specialized agents that delegate and act in co...
Effective model-based reinforcement learning in stochastic environments requires planning that accounts for predictiv...
Fairness evaluation concerns not only what a model produces, but also what its outputs ought to be compared against. ...
Cognitive AI seeks to move beyond language generation and autonomous task execution toward systems capable of sustain...
Retrieval-augmented generation (RAG) imposes a prefill cost proportional to retrieved context length, and -- with Tra...
The efficiency of a datacenter rests on its control plane policies. Designing these policies is increasingly hard: th...
World Action Models (WAMs) augment robot policies with action-conditioned predicted futures, but a plausible future a...
Sparse retrieval underpins modern search systems, from web search to retrieval-augmented generation. Existing work ha...
Artificial intelligence (AI) is increasingly central to power and energy systems, supporting modeling, forecasting, o...