Tag
该论文提出两种学习得到的潜空间蛋白质语言——PLL（基于冻结ESM-2的4,096状态上下文字母表）和SLL（改进的GCP-VQVAE Lite结构分词器），并分别在其上预训练自回归Transformer（PLLM和SLLM），在蛋白质序列与结构生成上相比氨基酸级自回归模型表现出更好的计算扩展性、采样质量以及与AlphaFold2相比约1000倍的推理速度。
This paper proposes a two-stage physics-constrained autoregressive transformer model to predict driver stop-go decisions and longitudinal trajectories at signalized intersections, using real-world data to enhance traffic safety.
This paper introduces World-Language-Action (WLA) models, embodied foundation models that jointly predict textual subtasks, subgoal images, and robot actions from text, images, and robot states, achieving state-of-the-art multi-task and long-horizon learning in simulated and real-world environments.
PRISM is a decoder-only autoregressive transformer that solves the inverse problem of multilayer thin-film optical coating design by jointly predicting material selection and thickness, achieving state-of-the-art performance with significantly smaller models.
UniT is a unified feed-forward model for geometry perception using a Group Autoregressive Transformer that integrates multiple paradigms (online/offline, multi-modal, long-horizon) while maintaining metric-scale accuracy via scale-adaptive loss and queue-style KV caching. It achieves state-of-the-art performance on ten benchmarks spanning seven tasks.