Lattice AI Research

Research focus

Training Efficiency & Optimization (2)Architecture Design (Transformers, SSMs, MoE) (1)Inference & Quantization (1)RLHF & Preference Learning (1)

Frequent co-authors

Yudong Wang (2)Di Zhang (1)Xun Wu (1)Shaohan Huang (1)

Papers (2)

Mar 5, 2026

Microsoft Research6d ago·also BIT, PKU

Sparse-BitNet: 1.58-bit LLMs are Naturally Friendly to Semi-Structured Sparsity

1.58-bit LLMs are surprisingly more resilient to sparsity than their full-precision counterparts, opening new avenues for extreme compression.

Di Zhang, Xun Wu, Shaohan Huang +9

Architecture Design (Transformers, SSMs, MoE)Inference & Quantization Training Efficiency & Optimization

Feb 26, 2026

1w ago·also Microsoft Research

Towards Better RL Training Data Utilization via Second-Order Rollout

LLMs trained with a novel "second-order rollout" that generates critiques in addition to responses learn more effectively from the same data, unlocking better reasoning.

Zhe Yang, Yudong Wang, Rang Li +2

RLHF & Preference Learning Training Efficiency & Optimization

Zhifang Sui

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (2)

Search

Zhifang Sui

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (2)