Yishan Li
Yishan Li has 3 papers in our library, in 2023–2026, most of them on Text generation; often with Bill Yuchen Lin, Biyuan Lin, Dingling Xu.
OpenAlex · ORCID · h-index 19 · 119 works · 1,380 citations OpenAlex, 2026-10-14
📄 3 works in our library · 🔗 Cited 84 times in our library · ⭐ Paper Stars ranking: #7046
Affiliations
🎓Sun Yat-sen UniversityGuangzhou ·
China2022–2026 · 4 papers🎓China University of Geosciences (Beijing)Beijing ·
China2026 · 1 paper🎓Nanjing University of Aeronautics and AstronauticsNanjing ·
China2026 · 1 paper🔬State Key Laboratory of Geological Processes and Mineral ResourcesWuhan2026 · 1 paper🎓Kaohsiung Medical UniversityKaohsiung ·
Taiwan2018–2024 · 7 papers🔬Genomics Research Center, Academia SinicaTaipei ·
Taiwan2006–2022 · 7 papers🔬Academia SinicaTaipei ·
Taiwan2006–2022 · 6 papers🎓National Yang Ming Chiao Tung UniversityHsinchu ·
Taiwan2007–2021 · 6 papers🎓Chongqing UniversityChongqing ·
China2016–2017 · 6 papers🎓Minnan Normal UniversityZhangzhou ·
China2024–2025 · 4 papers
Fields, topics and concepts
📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Arts and Humanities · 3 worksComputer Science · 3 worksArtificial Intelligence · 3 worksLanguage and Linguistics · 3 worksComputer Vision and Pattern Recognition · 2 worksInformation Systems · 1 work
🧭 Topics 1Research subjects and languages the work is about.Text generation · 1 work
Works by year
2020s
Papers
- TIGERScore: Towards Building Explainable Metric for All Text Generation Tasks (2023 · 50 citations)
Automatic metrics for text generation usually need gold references, work only in narrow domains such as machine translation, or give a bare number with no explanation. - RAGEval: Scenario Specific RAG Evaluation Dataset Generation Framework (2025 · 34 citations)
Testing RAG systems in specialised fields such as law, finance or medicine is hard. Building expert-annotated, domain-specific test sets is expensive and raises privacy issues, and lexical metrics like BLEU or ROUGE-L judge long answers poorly. RAGEval addresses this by generating evaluation data automatically.… - MiniCPM-SALA: Hybridizing Sparse and Linear Attention for Efficient Long-Context Modeling (2026 · 2 citations)
Full attention in Transformers is costly for very long inputs. Compute grows quadratically with length, and the KV-Cache can reach tens or hundreds of gigabytes at million-token scale. Sparse attention cuts computation but still keeps the whole cache. Linear attention is cheap but compresses context lossily and loses…
Co-authors here
- Bill Yuchen Lin
- Biyuan Lin
- Dingling Xu
- Dongfu Jiang
- Ge Zhang
- Kunlun Zhu
- MiniCPM Team
- Nan Zhang
- Ruobing Wang
- Shi Yu
- Shuo Wang
- Wenhao Huang
Venues
Cited by
Updated
This page in Markdown → https://md.research.elamur.ai/free/en/authors/yishan-li/