--- title: "Haodong Duan — author profile and papers" description: "Haodong Duan, Tongji University: 4 papers on Elamur Research (2023–2024), h-index 18 — affiliations, topics, co-authors and who cites them in our library." canonical: "https://research.elamur.ai/free/en/authors/haodong-duan/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/haodong-duan/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Haodong Duan # Haodong Duan Haodong Duan works on large language models and vision-language models, with a steady focus on how such systems should be evaluated. Their papers here, from 2023 and 2024, ask whether existing tests really measure the abilities they claim to, and they pair benchmark design with technical reports on open model families. The methods combine large-scale model training with careful benchmark construction: curated multiple-choice question sets, bilingual test versions, and probing models with key inputs withheld. The most cited work here is [MMBench](https://md.research.elamur.ai/free/en/papers/liu-2023-mmbench-multi-modal-model-all/) (2023). It is an objective benchmark of 3,217 multiple-choice questions spanning 20 fine-grained abilities under perception and reasoning, and it comes in parallel English and Chinese versions so that correct answers are not penalised for their wording. The [InternLM2 Technical Report](https://md.research.elamur.ai/free/en/papers/cai-2024-internlm2-technical-report/) (2024) documents open models at 1.8B, 7B and 20B parameters. It explains data preparation in detail, extends context length in stages, and introduces an alignment method called COOL RLHF. [Are We on the Right Way for Evaluating Large Vision-Language Models?](https://md.research.elamur.ai/free/en/papers/chen-2024-right-way-evaluating-large-vision/) (2024) shows that many items on popular multi-modal benchmarks can be answered without the image, through cues in the question, stored knowledge or apparent memorisation. In response it offers MMStar, a set of 1,500 items. The newest papers, all from 2024, extend the same lines of evaluation and open modelling. [InternLM-XComposer2](https://md.research.elamur.ai/free/en/papers/dong-2024-internlm-xcomposer2-mastering-free/) moves toward generation. It is a vision-language model that interprets images and also composes long articles interleaving text and pictures. It can be guided by outlines, written requirements and reference images rather than by a title alone. Duan's papers most often list the [Shanghai Artificial Intelligence Laboratory](https://md.research.elamur.ai/free/en/labs/shanghai-artificial-intelligence-laboratory/) (from 2022) and the [Chinese University of Hong Kong](https://md.research.elamur.ai/free/en/universities/chinese-university-of-hong-kong/) (from 2019), and some recent ones also list [Fudan University](https://md.research.elamur.ai/free/en/universities/fudan-university/). Frequent co-authors on the papers here include [Xiaoyi Dong](https://md.research.elamur.ai/free/en/authors/xiaoyi-dong/), [Dahua Lin](https://md.research.elamur.ai/free/en/authors/dahua-lin/) and [Kai Chen](https://md.research.elamur.ai/free/en/authors/kai-chen/). The work falls mainly under [Vision & language](https://md.research.elamur.ai/free/en/topics/multimodal/), with further papers on language model pretraining and on evaluation and benchmarks. Haodong Duan has 4 papers in our library, in 2023–2024, most of them on [Vision & language](https://md.research.elamur.ai/free/en/topics/multimodal/), [Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/), [Evaluation & benchmarks](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/); often with [Xiaoyi Dong](https://md.research.elamur.ai/free/en/authors/xiaoyi-dong/), [Dahua Lin](https://md.research.elamur.ai/free/en/authors/dahua-lin/), [Jiaqi Wang](https://md.research.elamur.ai/free/en/authors/jiaqi-wang/). 4 papers in our library · 2023–2024 [OpenAlex](https://openalex.org/A5028468431) · [ORCID](https://orcid.org/0000-0002-3052-4177) · [Wikidata](https://www.wikidata.org/wiki/Q131177237) · [**h-index 18**](https://openalex.org/authors/A5028468431) · [78 works](https://openalex.org/works?filter=author.id:A5028468431) · [2,269 citations](https://openalex.org/works?filter=author.id:A5028468431&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [📄 4 works in our library](https://md.research.elamur.ai/free/en/authors/haodong-duan/#papers) · [🔗 1,140 citations in our library](https://md.research.elamur.ai/free/en/authors/haodong-duan/cited-by/) · [6 more in our library without a page yet](#also) · [⭐ Paper Star rank #1464](https://md.research.elamur.ai/free/en/paper-stars/page/15/#star-1464) 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=Haodong+Duan) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Haodong+Duan) · [dblp](https://dblp.org/search?q=Haodong+Duan) · [ACL Anthology](https://aclanthology.org/search/?q=Haodong+Duan) ## Affiliations [🔬**Shanghai Artificial Intelligence Laboratory** Shanghai · China 2022–2026 · 33 papers*current*](https://md.research.elamur.ai/free/en/labs/shanghai-artificial-intelligence-laboratory/) · [🎓**Chinese University of Hong Kong** Hong Kong · Hong Kong 2019–2026 · 16 papers*current*](https://md.research.elamur.ai/free/en/universities/chinese-university-of-hong-kong/) · [🎓**Tongji University** Shanghai · China 2026 · 6 papers*current*](https://ror.org/03rc6as71) · [🎓**Fudan University** Shanghai · China 2024–2026 · 5 papers*current*](https://md.research.elamur.ai/free/en/universities/fudan-university/) · [🎓**Zhejiang University** Hangzhou · China 2026 · 3 papers*current*](https://md.research.elamur.ai/free/en/universities/zhejiang-university/) · [🏢**ByteDance** Beijing · China 2026 · 2 papers*current*](https://md.research.elamur.ai/free/en/companies/bytedance/) · [🎓**University of Michigan** Ann Arbor · United States 2026 · 1 paper*current*](https://md.research.elamur.ai/free/en/universities/university-of-michigan/) From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 1 more institution on single bylines are not listed. ## Topics [Vision & language (3)](https://md.research.elamur.ai/free/en/topics/multimodal/) · [Language models & pretraining (1)](https://md.research.elamur.ai/free/en/topics/language-models/) · [Evaluation & benchmarks (1)](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/) ## Works by year ### 2020s [**2024** 3 papers](https://md.research.elamur.ai/free/en/authors/haodong-duan/year/2024/) · [**2023** 1 paper](https://md.research.elamur.ai/free/en/papers/liu-2023-mmbench-multi-modal-model-all/) ## Papers The most cited first (OpenAlex and our library). 1. **[MMBench: Is Your Multi-modal Model an All-around Player?](https://md.research.elamur.ai/free/en/papers/liu-2023-mmbench-multi-modal-model-all/)** *(2023 · 907 citations)* Evaluating large vision-language models (VLMs) is hard to do well. Classic datasets such as VQAv2 or COCO Caption score answers by exact string match, so they penalise correct answers that use different wording, and they say little about specific skills. Human-judged evaluations such as OwlEval are small, costly and… 2. **[InternLM2 Technical Report](https://md.research.elamur.ai/free/en/papers/cai-2024-internlm2-technical-report/)** *(2024 · 299 citations)* Open-source LLMs still trail proprietary systems such as GPT-4, and earlier technical reports rarely explain how pre-training data was prepared. 3. **[Are We on the Right Way for Evaluating Large Vision-Language Models?](https://md.research.elamur.ai/free/en/papers/chen-2024-right-way-evaluating-large-vision/)** *(2024 · 263 citations)* Multi-modal benchmarks are supposed to measure how well large vision-language models (LVLMs) use images. The authors show that many existing test items can be solved without the picture. Some questions give away the answer in their wording, and others can be answered from general knowledge stored in the language… 4. **[InternLM-XComposer2: Mastering Free-form Text-Image Composition and Comprehension in Vision-Language Large Model](https://md.research.elamur.ai/free/en/papers/dong-2024-internlm-xcomposer2-mastering-free/)** *(2024 · 92 citations)* InternLM-XComposer2 is a vision-language model that both understands images and writes long documents mixing text and pictures. Its predecessor could only build such articles from a title. The new model accepts much richer inputs: outlines, detailed written requirements and reference images. ## Also in our library 6 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 19 co-authors with a page, the most papers together first. - [Xiaoyi Dong](https://md.research.elamur.ai/free/en/authors/xiaoyi-dong/) ×3 - [Dahua Lin](https://md.research.elamur.ai/free/en/authors/dahua-lin/) ×2 - [Jiaqi Wang](https://md.research.elamur.ai/free/en/authors/jiaqi-wang/) ×2 - [Kai Chen](https://md.research.elamur.ai/free/en/authors/kai-chen/) ×2 - [Maosong Cao](https://md.research.elamur.ai/free/en/authors/maosong-cao/) ×2 - [Pan Zhang](https://md.research.elamur.ai/free/en/authors/pan-zhang/) ×2 - [Songyang Zhang](https://md.research.elamur.ai/free/en/authors/songyang-zhang/) ×2 - [Yuhang Zang](https://md.research.elamur.ai/free/en/authors/yuhang-zang/) ×2 - [Zehui Chen](https://md.research.elamur.ai/free/en/authors/zehui-chen/) ×2 - [Bin Wang](https://md.research.elamur.ai/free/en/authors/bin-wang/) - [Bo Li](https://md.research.elamur.ai/free/en/authors/bo-li/) - [Conghui He](https://md.research.elamur.ai/free/en/authors/conghui-he/) ## Cited by Papers of our library cite their works 1,140 times (of the 140,884 papers whose bibliographies we read); [64 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/haodong-duan/cited-by/).