--- title: "Zhengyuan Yang — author profile and papers" description: "Zhengyuan Yang, Microsoft (United States): 3 papers on Elamur Research (2022–2023), h-index 29 — affiliations, topics, co-authors and who cites them in…" canonical: "https://research.elamur.ai/free/en/authors/zhengyuan-yang/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/zhengyuan-yang/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Zhengyuan Yang # Zhengyuan Yang Zhengyuan Yang works on [vision and language](https://md.research.elamur.ai/free/en/topics/multimodal/) and on [language models](https://md.research.elamur.ai/free/en/topics/language-models/), with a steady interest in how to evaluate what large models can actually do. The papers here, from 2022 and 2023, probe the abilities and the trustworthiness of large multimodal models and GPT-3-style systems. They do so through hand-built test cases, benchmarks assembled from combinations of skills, and prompting experiments. The most cited of these is [MM-Vet](https://md.research.elamur.ai/free/en/papers/yu-2023-mm-vet-evaluating-large-multimodal/) (2023). It is a benchmark that treats hard multimodal tasks as blends of six basic abilities: recognition, reading text in images, world knowledge, language generation, spatial awareness and math. It spans 16 combinations of these abilities, and because answers can be open-ended, it uses GPT-4 with a few-shot prompt to grade them. [The Dawn of LMMs](https://md.research.elamur.ai/free/en/papers/yang-2023-dawn-lmms-preliminary-explorations-gpt/) (2023) takes a qualitative look at GPT-4V(ision). It reports no benchmark numbers and instead works through curated examples in areas such as charts and documents, medical images, multilingual input, video, and IQ-style puzzles. It also adds notes on accepted input formats and useful prompting techniques. Earlier, [Prompting GPT-3 To Be Reliable](https://md.research.elamur.ai/free/en/papers/si-2022-prompting-gpt-3-reliable/) (2022) broke reliability into four parts: factuality, calibration, social bias, and how well a model generalizes. The paper tried simple prompting strategies on each part, mainly with code-davinci-002, and compared the results with smaller fine-tuned models such as BERT and T5. Across these papers, the newest work from 2023 shifts from text-only language models to models that combine images and text, and it keeps evaluation as the central question. Yang is currently at Microsoft in Redmond, and earlier bylines list the University of Rochester. Co-authors on the papers here include [Jianfeng Wang](https://md.research.elamur.ai/free/en/authors/jianfeng-wang/), [Lijuan Wang](https://md.research.elamur.ai/free/en/authors/lijuan-wang/), Kevin Lin, Linjie Li, Zicheng Liu and [Chenglei Si](https://md.research.elamur.ai/free/en/authors/chenglei-si/). 3 papers on this site · 2022–2023 current affiliation [Microsoft (United States)](https://md.research.elamur.ai/free/en/companies/microsoft/), Redmond, United States; [National University of Defense Technology](https://md.research.elamur.ai/free/en/universities/national-university-of-defense-technology/), Changsha, China; [Jilin University](https://md.research.elamur.ai/free/en/universities/jilin-university/), Changchun, China Zhengyuan Yang has 3 papers on this site, in 2022–2023, most of them on [Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/), [Vision & language](https://md.research.elamur.ai/free/en/topics/multimodal/), [Bias, fairness & safety](https://md.research.elamur.ai/free/en/topics/bias-fairness-safety/); often with [Jianfeng Wang](https://md.research.elamur.ai/free/en/authors/jianfeng-wang/), [Lijuan Wang](https://md.research.elamur.ai/free/en/authors/lijuan-wang/), [Kevin Lin](https://md.research.elamur.ai/free/en/authors/kevin-lin/). [OpenAlex](https://openalex.org/A5050209478) · [ORCID](https://orcid.org/0000-0002-5808-0889) · [**h-index 29**](https://openalex.org/authors/A5050209478) · [106 works](https://openalex.org/works?filter=author.id:A5050209478) · [4 442 citations](https://openalex.org/works?filter=author.id:A5050209478&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [3 works on this site](https://md.research.elamur.ai/free/en/authors/zhengyuan-yang/#papers) · [837 citations in our library](https://md.research.elamur.ai/free/en/authors/zhengyuan-yang/cited-by/) · 5 more works in our library 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=Zhengyuan+Yang) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Zhengyuan+Yang) · [dblp](https://dblp.org/search?q=Zhengyuan+Yang) · [ACL Anthology](https://aclanthology.org/search/?q=Zhengyuan+Yang) ## Affiliations - [Microsoft (United States)](https://md.research.elamur.ai/free/en/companies/microsoft/) · Redmond, United States —current · 2021–2026 on 40 papers - [Jilin University](https://md.research.elamur.ai/free/en/universities/jilin-university/) · Changchun, China —current · 2024–2026 on 4 papers - [National University of Defense Technology](https://md.research.elamur.ai/free/en/universities/national-university-of-defense-technology/) · Changsha, China —current · 2025–2026 on 2 papers - [Southwest Jiaotong University](https://ror.org/00hn7w693) · Chengdu, China —2023–2025 on 2 papers - [Zhejiang Normal University](https://ror.org/01vevwk45) · Jinhua, China —2021–2025 on 2 papers - [University of Rochester](https://md.research.elamur.ai/free/en/universities/university-of-rochester/) · Rochester, United States —2017–2024 on 20 papers From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 9 more institutions on single bylines are not listed. ## Topics [Language models & pretraining 2](https://md.research.elamur.ai/free/en/topics/language-models/) · [Vision & language 2](https://md.research.elamur.ai/free/en/topics/multimodal/) · [Bias, fairness & safety 1](https://md.research.elamur.ai/free/en/topics/bias-fairness-safety/) ## Works by year ### 2020s [**2023** 2 papers](https://md.research.elamur.ai/free/en/authors/zhengyuan-yang/year/2023/) · [**2022** 1 paper](https://md.research.elamur.ai/free/en/papers/si-2022-prompting-gpt-3-reliable/) ## Papers The most cited first (OpenAlex and our library). 1. **[MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities](https://md.research.elamur.ai/free/en/papers/yu-2023-mm-vet-evaluating-large-multimodal/)** *(2023 · 328 citations)* Classic vision-language benchmarks such as VQA v2, COCO captioning, and TextVQA each probe only one or two skills. 2. **[The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)](https://md.research.elamur.ai/free/en/papers/yang-2023-dawn-lmms-preliminary-explorations-gpt/)** *(2023 · 303 citations)* Microsoft researchers present an early, mostly qualitative study of GPT-4V(ision), the vision-enabled version of GPT-4 from OpenAI. 3. **[Prompting GPT-3 To Be Reliable](https://md.research.elamur.ai/free/en/papers/si-2022-prompting-gpt-3-reliable/)** *(2022 · 206 citations)* The paper asks how far plain prompting can make GPT-3 trustworthy, beyond raw accuracy. The authors split reliability into four facets drawn from the ML-safety literature: generalizability, social bias, calibration and factuality. For each facet they design simple prompting strategies and test them mainly with… ## Also in our library 5 more works of theirs are in our library without a page here yet; they will appear as we publish them. ## Co-authors here 11 co-authors with a page, the most papers together first. - [Jianfeng Wang](https://md.research.elamur.ai/free/en/authors/jianfeng-wang/) ×3 - [Lijuan Wang](https://md.research.elamur.ai/free/en/authors/lijuan-wang/) ×3 - [Kevin Lin](https://md.research.elamur.ai/free/en/authors/kevin-lin/) ×2 - [Linjie Li](https://md.research.elamur.ai/free/en/authors/linjie-li/) ×2 - [Zicheng Liu](https://md.research.elamur.ai/free/en/authors/zicheng-liu/) ×2 - [Chenglei Si](https://md.research.elamur.ai/free/en/authors/chenglei-si/) - [Jordan Boyd-Graber](https://md.research.elamur.ai/free/en/authors/jordan-boyd-graber/) - [Shuohang Wang](https://md.research.elamur.ai/free/en/authors/shuohang-wang/) - [Weihao Yu](https://md.research.elamur.ai/free/en/authors/weihao-yu/) - [Xinchao Wang](https://md.research.elamur.ai/free/en/authors/xinchao-wang/) - [Zhe Gan](https://md.research.elamur.ai/free/en/authors/zhe-gan/) ## Cited by Papers of our library cite their works 837 times (of the 140 884 papers whose bibliographies we read); [81 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/zhengyuan-yang/cited-by/).