Peng CHENG

Training Matryoshka Mixture-of-Experts for Elastic Inference-Time Expert Utilization

Recycling Pretrained Checkpoints: Orthogonal Growth of Mixture-of-Experts for Efficient Large Language Model Pre-Training

Learning from the Best, Differently: A Diversity-Driven Rethinking on Data Selection

Retrieval as Reasoning: Learning to Select and Generate with LLMs

Automated Proof Generation for Rust Code via Self-Evolution

Optimizing Large Language Model Training Using FP4 Quantization

Integrative Decoding: Improving Factuality via Implicit Self-consistency