--- title: "Shirong Ma — author profile and papers" description: "Shirong Ma, Nanjing University of Information Science and Technology: 3 papers on Elamur Research (2022–2025), h-index 12 — affiliations, topics…" canonical: "https://research.elamur.ai/free/en/authors/shirong-ma/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/shirong-ma/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Shirong Ma # Shirong Ma Shirong Ma works on natural language processing, with papers here that span Chinese text correction and large language models for code and reasoning. A recurring concern is the shortage of suitable training signal: hand-annotated data for correcting grammatical errors in Chinese, code and mathematics corpora for pretraining, and human-written reasoning traces for teaching models to think step by step. The methods range from rule-based generation of synthetic training examples to large-scale continued pretraining and reinforcement learning. The topics include [Grammatical error correction & learner language](https://md.research.elamur.ai/free/en/topics/grammatical-error-correction/), [Code & programming languages](https://md.research.elamur.ai/free/en/topics/code-and-programming/) and [Reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/). The earliest of these papers, [Linguistic Rules-Based Corpus Generation for Native Chinese Grammatical Error Correction](https://md.research.elamur.ai/free/en/papers/ma-2022-linguistic-rules-based-corpus-generation/) (2022), argues that existing Chinese benchmarks mostly reflect the obvious mistakes of second-language learners. It proposes CLG, which inserts errors into clean sentences using rules drawn from Chinese linguistics references, and releases NaCGEC, a test set aimed at the subtler errors native speakers make. [DeepSeek-Coder-V2](https://md.research.elamur.ai/free/en/papers/deepseek-2024-ai-deepseek-coder-v2-breaking-barrier/) (2024) extends a Mixture-of-Experts checkpoint with trillions of further tokens weighted toward code and mathematics. It covers hundreds of programming languages and a 128K-token context window, with the aim of narrowing the gap with closed-source code models. The most recent work, [DeepSeek-R1](https://md.research.elamur.ai/free/en/papers/deepseek-2025-ai-deepseek-r1-incentivizing-reasoning-capability/) (2025), moves from data construction toward training methods for reasoning. It starts from a base model with no supervised fine-tuning and applies GRPO. The rewards are defined by simple rules that verify whether the model's answer to a mathematics, programming or logic problem is correct. The resulting model learns on its own to produce longer chains of reasoning, verify its steps and reflect on its answers. It is the most cited of Ma's papers here, with 4593 citations. Ma's current affiliation is [DeepSeek](https://md.research.elamur.ai/free/en/companies/deepseek/) in Hangzhou. Earlier papers from 2022 to 2025 list [Tsinghua University](https://md.research.elamur.ai/free/en/universities/tsinghua-university/) and its Shenzhen International Graduate School. Frequent co-authors include [Daya Guo](https://md.research.elamur.ai/free/en/authors/daya-guo/), [Peiyi Wang](https://md.research.elamur.ai/free/en/authors/peiyi-wang/) and [Runxin Xu](https://md.research.elamur.ai/free/en/authors/runxin-xu/), along with the [DeepSeek-AI](https://md.research.elamur.ai/free/en/authors/deepseek-ai/) team. 3 papers on this site · 2022–2025 current affiliation [Nanjing University of Information Science and Technology](https://ror.org/02y0rxk19), Nanjing, China; [DeepSeek (China)](https://md.research.elamur.ai/free/en/companies/deepseek/), Hangzhou, China Shirong Ma has 3 papers on this site, in 2022–2025, most of them on [Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/), [Reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/), [Code & programming languages](https://md.research.elamur.ai/free/en/topics/code-and-programming/); often with [Daya Guo](https://md.research.elamur.ai/free/en/authors/daya-guo/), [DeepSeek-AI](https://md.research.elamur.ai/free/en/authors/deepseek-ai/), [Dejian Yang](https://md.research.elamur.ai/free/en/authors/dejian-yang/). [OpenAlex](https://openalex.org/A5057030499) · [ORCID](https://orcid.org/0000-0002-0035-9078) · [**h-index 12**](https://openalex.org/authors/A5057030499) · [45 works](https://openalex.org/works?filter=author.id:A5057030499) · [2 323 citations](https://openalex.org/works?filter=author.id:A5057030499&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [3 works on this site](https://md.research.elamur.ai/free/en/authors/shirong-ma/#papers) · [4 321 citations in our library](https://md.research.elamur.ai/free/en/authors/shirong-ma/cited-by/) · 3 more works in our library · [⭐ Paper Star rank #526](https://md.research.elamur.ai/free/en/paper-stars/page/6/#star-526) The profile is the library's match of the name; two people with one name are not always told apart. Elsewhere: [Google Scholar](https://scholar.google.com/citations?view_op=search_authors&mauthors=Shirong+Ma) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Shirong+Ma) · [dblp](https://dblp.org/search?q=Shirong+Ma) · [ACL Anthology](https://aclanthology.org/search/?q=Shirong+Ma) ## Affiliations - [DeepSeek (China)](https://md.research.elamur.ai/free/en/companies/deepseek/) · Hangzhou, China —current · 2025–2026 on 3 papers - [Nanjing University of Information Science and Technology](https://ror.org/02y0rxk19) · Nanjing, China —current · 2026 on 1 paper - [Tsinghua University](https://md.research.elamur.ai/free/en/universities/tsinghua-university/) · Beijing, China —2022–2025 on 15 papers - [Tsinghua Shenzhen International Graduate School](https://md.research.elamur.ai/free/en/universities/tsinghua-shenzhen-international-graduate-school/) · Shenzhen, China —2022–2025 on 14 papers - [Air Force Medical University](https://ror.org/00ms48f15) · Xi'an, China —2007–2025 on 11 papers - [Xijing Hospital](https://ror.org/05cqe9350) · Xi’an, China —2007–2025 on 5 papers From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 3 more institutions on single bylines are not listed. ## Topics [Language models & pretraining 1](https://md.research.elamur.ai/free/en/topics/language-models/) · [Reasoning 1](https://md.research.elamur.ai/free/en/topics/reasoning/) · [Code & programming languages 1](https://md.research.elamur.ai/free/en/topics/code-and-programming/) · [Grammatical error correction & learner language 1](https://md.research.elamur.ai/free/en/topics/grammatical-error-correction/) ## Works by year ### 2020s [**2025** 1 paper](https://md.research.elamur.ai/free/en/papers/deepseek-2025-ai-deepseek-r1-incentivizing-reasoning-capability/) · [**2024** 1 paper](https://md.research.elamur.ai/free/en/papers/deepseek-2024-ai-deepseek-coder-v2-breaking-barrier/) · [**2022** 1 paper](https://md.research.elamur.ai/free/en/papers/ma-2022-linguistic-rules-based-corpus-generation/) ## Papers The most cited first (OpenAlex and our library). 1. **[DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://md.research.elamur.ai/free/en/papers/deepseek-2025-ai-deepseek-r1-incentivizing-reasoning-capability/)** *(2025 · 4 593 citations)* Chain-of-thought prompting and supervised fine-tuning have made LLMs better at reasoning, but both rely on reasoning traces written by humans. 2. **[DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence](https://md.research.elamur.ai/free/en/papers/deepseek-2024-ai-deepseek-coder-v2-breaking-barrier/)** *(2024 · 195 citations)* Open-source code models have lagged behind closed-source offerings, with GPT4-Turbo, Gemini 1.5 Pro and Claude 3 Opus among the leaders. 3. **[Linguistic Rules-Based Corpus Generation for Native Chinese Grammatical Error Correction](https://md.research.elamur.ai/free/en/papers/ma-2022-linguistic-rules-based-corpus-generation/)** *(2022 · 59 citations)* Most Chinese Grammatical Error Correction (CGEC) research faces two problems: there is little high-quality annotated training data, and standard test sets such as NLPCC and CGED contain errors from L2 learners. ## Also in our library 3 more works of theirs are in our library without a page here yet; they will appear as we publish them. ## Co-authors here 12 of 17 co-authors with a page, the most papers together first. - [Daya Guo](https://md.research.elamur.ai/free/en/authors/daya-guo/) ×2 - [DeepSeek-AI](https://md.research.elamur.ai/free/en/authors/deepseek-ai/) ×2 - [Dejian Yang](https://md.research.elamur.ai/free/en/authors/dejian-yang/) ×2 - [Peiyi Wang](https://md.research.elamur.ai/free/en/authors/peiyi-wang/) ×2 - [Qihao Zhu](https://md.research.elamur.ai/free/en/authors/qihao-zhu/) ×2 - [Runxin Xu](https://md.research.elamur.ai/free/en/authors/runxin-xu/) ×2 - [Haitao Zheng](https://md.research.elamur.ai/free/en/authors/haitao-zheng/) - [Haowei Zhang](https://md.research.elamur.ai/free/en/authors/haowei-zhang/) - [Huazuo Gao](https://md.research.elamur.ai/free/en/authors/huazuo-gao/) - [Junxiao Song](https://md.research.elamur.ai/free/en/authors/junxiao-song/) - [Qingyu Zhou](https://md.research.elamur.ai/free/en/authors/qingyu-zhou/) - [Shulin Huang](https://md.research.elamur.ai/free/en/authors/shulin-huang/) ## Cited by Papers of our library cite their works 4 321 times (of the 140 884 papers whose bibliographies we read); [98 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/shirong-ma/cited-by/).