Profiles2025-2026
Xu Han
Department of Computer Science and Technology, Tsinghua University, Beijing
9Papers
1Last author
5Unique co-authors
0First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers9
- InfLLM-V2: Dense-Sparse Switchable Attention for Seamless Short-to-Long AdaptationAuthor 12 of 13ICLR 2026 · 32 citations
- UniSVQ: 2-bit Unified Scalar-Vector QuantizationAuthor 5 of 7ICML 2026 · 2 citations
- LC-QAT: Data-Efficient 2-Bit QAT for LLMs via Linear-Constrained Vector QuantizationAuthor 5 of 5ICML 2026 · 1 citation
- APB-V: Accelerating Long-Video Understanding via Sequence-Parallelism-aware Approximate AttentionAuthor 3 of 10ACL 2026
- FR-Spec: Accelerating Large-Vocabulary Language Models via Frequency-Ranked Speculative SamplingAuthor 3 of 12ACL 2025 · 14 citations
- BurstEngine: An efficient distributed framework for training transformers On extremely Long sequences of over 1M tokensAuthor 3 of 7SC 2025 · 1 citation
- APB: Accelerating Distributed Long-Context Inference by Passing Compressed Context Blocks across GPUsAuthor 3 of 10ACL 2025
- Cost-Optimal Grouped-Query Attention for Long-Context ModelingAuthor 6 of 8EMNLP 2025
- Sparsing Law: Towards Large Language Models with Greater Activation SparsityAuthor 3 of 10ICML 2025
