4 papers
Offline Semantic Guidance for Efficient Vision-Language-Action Policy Distillation
Jin Shi, Brady Zhang, Yishun Lu
Billion-parameter Vision-Language-Action (VLA) policies have recently shown impressive performance in robotic manipulation, yet their size and inference cost remain major obstacles…
Runtime-Orchestrated Second-Order Optimization for Scalable LLM Training
Yishun Lu, Junhao Zhang, Zeyu Yang +1
Second-order methods offer an attractive path toward more sample-efficient LLM training, but their practical use is often blocked by the systems cost of maintaining and updating la…
Second-Order Multi-Level Variance Correction for Modality Competition in Multimodal Models
Yishun Lu, Wes Armour
Autoregressive next-token training offers a unified formulation for image generation and text understanding, but it also creates strong modality competition that destabilizes optim…
Beyond the Mean: Fisher-Orthogonal Projection for Natural Gradient Descent in Large Batch Training
Yishun Lu, Wesley Armour
Modern GPUs are equipped with large amounts of high-bandwidth memory, enabling them to support mini-batch sizes of up to tens of thousands of training samples. However, most existi…