Tsinghua NLP Group · 8 records

dateidtitleauthorsvenue↑
2026-10-202610.26830Competing Learning in Hybrid ArchitecturesXingyu Shen et al.arXiv9
2026-10-162610.22194AutoSFT: An Agentic System for Automated Data Research in Supervised Fine-TuningBingxiang He et al.arXiv42
2026-10-152610.20582Latent Compaction with Test-Time TrainingYingfa Chen et al.arXiv5
2026-10-052609.39071LexReward: A Taxonomy-Driven Reward Framework for Legal Language ModelsYida Cai et al.arXiv29
2026-09-292609.33803Diffusion Reward ModelsXiangyang Wang et al.arXiv32
2026-09-092609.00787StudyBench: Can Self-Evolution Squeeze Textbooks for Olympiad Capability?Yinghao Chen et al.arXiv14
2026-08-182608.14441PACE-Bench: Benchmarking Physics Adaptation via Code Evolution in Dynamic EnvironmentsYuhao Zhan et al.arXiv29
2026-05-122605.10933DECO: Sparse Mixture-of-Experts with Dense-Comparable Performance on End-Side DevicesChenyang Song et al.arXiv2