3 papers
cs.LG2026
Turing LLM Post-Training: Calibrated Distillation and Efficient Dynamic-MoE Execution
Yue He, Jiahui Hu, Shaman Tang +19
We study post-training for dynamically routed mixture-of-experts (MoE) models, focusing on integrating independently optimized specialists while keeping consolidation efficient. Ou…
cs.AI2026
TuringLLM: Efficiently Scaling Foundation Models Toward Physical AI
Yuheng Zhang, Yizhao Wang, Da Zhu +19
We present Turing-20B-A2B, a 20B-parameter Mixture-of-Experts language model that activates approximately 2B parameters per token, designed for long-context and latency-sensitive p…
cs.CV2026
TuringViT: Making SOTA Vision Transformers Accessible to All
Qiman Wu, Hanlin Chen, Lyujie Chen +19
Modern VLMs and VLA systems commonly adopt off-the-shelf ViTs such as SigLIP2 as visual encoders, but diverse downstream requirements in latency, temporal modeling, and VLM integra…