--- title: "Xu Jiang — author profile and papers" description: "Xu Jiang, State Grid Jilin Electric Power: 2 papers on Elamur Research (2021–2022), h-index 12 — affiliations, topics, co-authors and who cites them in…" canonical: "https://research.elamur.ai/free/en/authors/xu-jiang/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/xu-jiang/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Xu Jiang # Xu Jiang Xu Jiang's papers here deal with large [language models](https://md.research.elamur.ai/free/en/topics/language-models/) and how to make their behaviour match what people actually ask for. The common thread is learning from human judgement: people write example answers or compare candidate outputs, a reward model is fitted to those preferences, and reinforcement learning then tunes the language model against it. Both works build on GPT-3 and appeared in 2021 and 2022. The most cited of them, [Training language models to follow instructions with human feedback](https://md.research.elamur.ai/free/en/papers/ouyang-2022-training-language-models-follow-instructions/), starts from the observation that scaling a model up does not by itself stop it from inventing facts, producing toxic text or ignoring the request. Hired labelers first wrote demonstration answers to prompts, which were used for supervised fine-tuning. They then ranked several model outputs, a reward model was trained on those rankings, and the policy was optimised with PPO. The earlier [WebGPT](https://md.research.elamur.ai/free/en/papers/nakano-2021-webgpt-browser-assisted-question-answering/) applies the same feedback-driven training to long-form [question answering](https://md.research.elamur.ai/free/en/topics/question-answering/). GPT-3 is given a simple text-only browser in which it can search, click links, scroll and quote passages, and the quoted passages become citations for its final answer. Training combined behaviour cloning on human browsing demonstrations, a reward model learned from pairwise comparisons of answers, and further reinforcement learning together with rejection sampling. Taken together, the two papers move from grounding answers in retrieved web sources to the broader goal of aligning a general-purpose model with user intent. Both were written by teams at OpenAI. Co-authors on both include [Long Ouyang](https://md.research.elamur.ai/free/en/authors/long-ouyang/), [Jacob Hilton](https://md.research.elamur.ai/free/en/authors/jacob-hilton/) and [Jeff Wu](https://md.research.elamur.ai/free/en/authors/jeff-wu/). Other collaborators include [Reiichiro Nakano](https://md.research.elamur.ai/free/en/authors/reiichiro-nakano/) and [John Schulman](https://md.research.elamur.ai/free/en/authors/john-schulman/). Xu Jiang has 2 papers in our library, in 2021–2022, most of them on [Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/), [Question answering](https://md.research.elamur.ai/free/en/topics/question-answering/); often with [Jacob Hilton](https://md.research.elamur.ai/free/en/authors/jacob-hilton/), [Jeff Wu](https://md.research.elamur.ai/free/en/authors/jeff-wu/), [Long Ouyang](https://md.research.elamur.ai/free/en/authors/long-ouyang/). 2 papers in our library · 2021–2022 current affiliation [State Grid Jilin Electric Power](https://ror.org/03fsc3516) · Changchun Shi China [OpenAlex](https://openalex.org/A5103934038) · [**h-index 12**](https://openalex.org/authors/A5103934038) · [36 works](https://openalex.org/works?filter=author.id:A5103934038) · [4,565 citations](https://openalex.org/works?filter=author.id:A5103934038&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [2 works in our library](https://md.research.elamur.ai/free/en/authors/xu-jiang/#papers) · [6,424 citations in our library](https://md.research.elamur.ai/free/en/authors/xu-jiang/cited-by/) · [⭐ Paper Star rank #390](https://md.research.elamur.ai/free/en/paper-stars/page/4/#star-390) 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=Xu+Jiang) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Xu+Jiang) · [dblp](https://dblp.org/search?q=Xu+Jiang) · [ACL Anthology](https://aclanthology.org/search/?q=Xu+Jiang) ## Affiliations [🏢**State Grid Jilin Electric Power** Changchun Shi · China 2025 · 1 paper*current*](https://ror.org/03fsc3516) · [🏛️**Chinese Academy of Agricultural Sciences** Beijing · China 2017–2024 · 10 papers](https://ror.org/0313jb750) · [🔬**Institute of Agricultural Resources and Regional Planning** Beijing · China 2017–2024 · 10 papers](https://ror.org/01nrzdp21) · [🎓**Shanghai Jiao Tong University** Shanghai · China 2017–2024 · 4 papers](https://md.research.elamur.ai/free/en/universities/shanghai-jiao-tong-university/) · [🔬**State Key Laboratory of Ocean Engineering** Shanghai 2023–2024 · 3 papers](https://ror.org/01m3vks35) From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 12 more institutions on single bylines are not listed. ## Topics [Language models & pretraining 1](https://md.research.elamur.ai/free/en/topics/language-models/) · [Question answering 1](https://md.research.elamur.ai/free/en/topics/question-answering/) ## Works by year ### 2020s [**2022** 1 paper](https://md.research.elamur.ai/free/en/papers/ouyang-2022-training-language-models-follow-instructions/) · [**2021** 1 paper](https://md.research.elamur.ai/free/en/papers/nakano-2021-webgpt-browser-assisted-question-answering/) ## Papers The most cited first (OpenAlex and our library). 1. **[Training language models to follow instructions with human feedback](https://md.research.elamur.ai/free/en/papers/ouyang-2022-training-language-models-follow-instructions/)** *(2022 · 5,770 citations)* Making language models larger does not by itself make them more helpful: they make up facts, generate toxic text, or disregard the user's request. 2. **[WebGPT: Browser-assisted question-answering with human feedback](https://md.research.elamur.ai/free/en/papers/nakano-2021-webgpt-browser-assisted-question-answering/)** *(2021 · 687 citations)* Long-form question answering (LFQA) asks a system to write a paragraph-length reply to an open-ended question, and such systems still fall short of human quality. ## Co-authors here 12 of 14 co-authors with a page, the most papers together first. - [Jacob Hilton](https://md.research.elamur.ai/free/en/authors/jacob-hilton/) ×2 - [Jeff Wu](https://md.research.elamur.ai/free/en/authors/jeff-wu/) ×2 - [Long Ouyang](https://md.research.elamur.ai/free/en/authors/long-ouyang/) ×2 - [Alex Ray](https://md.research.elamur.ai/free/en/authors/alex-ray/) - [Chong Zhang](https://md.research.elamur.ai/free/en/authors/chong-zhang/) - [Christopher Hesse](https://md.research.elamur.ai/free/en/authors/christopher-hesse/) - [Diogo Almeida](https://md.research.elamur.ai/free/en/authors/diogo-almeida/) - [John Schulman](https://md.research.elamur.ai/free/en/authors/john-schulman/) - [Karl Cobbe](https://md.research.elamur.ai/free/en/authors/karl-cobbe/) - [Pamela Mishkin](https://md.research.elamur.ai/free/en/authors/pamela-mishkin/) - [Reiichiro Nakano](https://md.research.elamur.ai/free/en/authors/reiichiro-nakano/) - [Sandhini Agarwal](https://md.research.elamur.ai/free/en/authors/sandhini-agarwal/) ## Cited by Papers of our library cite their works 6,424 times (of the 140,884 papers whose bibliographies we read); [494 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/xu-jiang/cited-by/).