History Replays Itself: Unleashing Long-Context Capability in Linear Attention
Paper • 2610.15715 • Published • 14
Liangyu Wang
ly4096
AI & ML interests
Efficient reinforcement learning (RL) for LLMs reasoning
Distributed training and inference of LLMs
Efficient algorithm and infrastructure design for LLMs
Recent Activity
upvoted a paper 3 days ago
History Replays Itself: Unleashing Long-Context Capability in Linear Attention upvoted a paper 9 days ago
Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight liked a model 2 months ago
Qwen/Qwen3.8-2.4T-A95B-FP8Organizations
upvoted a paper 3 days ago
upvoted a paper 9 days ago
liked a
model 2 months ago
upvoted a paper 5 months ago
authored a
paper 8 months ago
submitted a
paper to Daily Papers 8 months ago
upvoted 3
papers 8 months ago
Canzona: A Unified, Asynchronous, and Load-Balanced Framework for Distributed Matrix-based Optimizers
Paper • 2602.06079 • Published • 22
FlashDP: Private Training Large Language Models with Efficient DP-SGD
Paper • 2507.01154 • Published • 1
Infinite Sampling: Efficient and Stable Grouped RL Training for Large Language Models
Paper • 2506.22950 • Published • 1
authored 3
papers about 1 year ago
Infinite Sampling: Efficient and Stable Grouped RL Training for Large Language Models
Paper • 2506.22950 • Published • 1
FlashDP: Private Training Large Language Models with Efficient DP-SGD
Paper • 2507.01154 • Published • 1
DistZO2: High-Throughput and Memory-Efficient Zeroth-Order Fine-tuning LLMs with Distributed Parallel Computing
Paper • 2507.03211 • Published
upvoted a paper over 1 year ago