--- title: "Linfeng Song — author profile and papers" description: "Linfeng Song, Tencent (China): 6 papers on Elamur Research (2017–2024), h-index 23 — affiliations, topics, co-authors and who cites them in our library." canonical: "https://research.elamur.ai/free/en/authors/linfeng-song/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/linfeng-song/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Linfeng Song # Linfeng Song Linfeng Song works in natural language processing, with much of the work here built around Abstract Meaning Representation (AMR), a graph-based encoding of sentence meaning. Recurring questions are how such graphs can be turned back into fluent text and whether they help machine translation. A second line deals with pronouns that Chinese leaves unstated, and the newest paper looks at the efficiency of reasoning in large language models. The methods range from grammar-based generation with beam search to graph neural encoders, multi-task models on BERT, and purpose-built test sets. The most cited paper here is [A Graph-to-Sequence Model for AMR-to-Text Generation](https://md.research.elamur.ai/free/en/papers/song-2018-graph-sequence-model-amr-text/) (2018). It stops flattening the graph into a token string and instead uses a graph-state LSTM in which every node keeps its own state and exchanges information with its neighbours along labelled edges. An earlier 2017 paper reached the same goal without neural networks: it learned a synchronous node replacement grammar from aligned graph–sentence pairs and scored 25.62 BLEU on LDC2015E86. [Semantic Neural Machine Translation Using AMR](https://md.research.elamur.ai/free/en/papers/song-2019-semantic-neural-machine-translation-amr/) (2019) added a graph recurrent encoder of source-side AMR to an English–German translator. The gain over a strong attention baseline was larger when training data was reduced. The more recent papers turn to Chinese zero pronouns and to reasoning models. ZPR2 (2020) recovers and resolves dropped pronouns together in a single multi-task model, without relying on gold syntax. [GuoFeng](https://md.research.elamur.ai/free/en/papers/xu-2022-guofeng-benchmark-zero-pronoun-recovery/) (2022) supplies a human-annotated Chinese–English benchmark for translating such pronouns. [Do NOT Think That Much for 2+3=?](https://md.research.elamur.ai/free/en/papers/chen-2024-think-that-much-2-3/) (2024) shows that o1-style models keep producing redundant solution rounds on easy problems after the answer is already found. It proposes efficiency measures to report alongside accuracy. Song has been at [Tencent](https://md.research.elamur.ai/free/en/companies/tencent/) in Shenzhen since 2019 and is also affiliated with the [University of Hong Kong](https://md.research.elamur.ai/free/en/universities/university-of-hong-kong/). Song earned a Ph.D. in computer science at the [University of Rochester](https://md.research.elamur.ai/free/en/universities/university-of-rochester/) (2014–2019), after earlier papers with the [Institute of Computing Technology](https://md.research.elamur.ai/free/en/labs/institute-of-computing-technology/), [Chinese Academy of Sciences](https://md.research.elamur.ai/free/en/government/chinese-academy-of-sciences/). Frequent co-authors include [Yue Zhang](https://md.research.elamur.ai/free/en/authors/yue-zhang/), [Daniel Gildea](https://md.research.elamur.ai/free/en/authors/daniel-gildea/) and [Zhiguo Wang](https://md.research.elamur.ai/free/en/authors/zhiguo-wang/). The papers here sit mainly under [semantic parsing and meaning representation](https://md.research.elamur.ai/free/en/topics/semantic-parsing/) and [machine translation](https://md.research.elamur.ai/free/en/topics/machine-translation/). 6 papers on this site · 2017–2024 current affiliation [Tencent (China)](https://md.research.elamur.ai/free/en/companies/tencent/), Shenzhen, China; [University of Hong Kong](https://md.research.elamur.ai/free/en/universities/university-of-hong-kong/), Pok Fu Lam, Hong Kong Linfeng Song has 6 papers on this site, in 2017–2024, most of them on [Semantic parsing & meaning representation](https://md.research.elamur.ai/free/en/topics/semantic-parsing/), [Machine translation](https://md.research.elamur.ai/free/en/topics/machine-translation/), [Text generation](https://md.research.elamur.ai/free/en/topics/text-generation/); often with [Yue Zhang](https://md.research.elamur.ai/free/en/authors/yue-zhang/), [Daniel Gildea](https://md.research.elamur.ai/free/en/authors/daniel-gildea/), [Zhiguo Wang](https://md.research.elamur.ai/free/en/authors/zhiguo-wang/). 2 of them are among the papers [Mova](https://md.research.elamur.ai/free/en/projects/mova/) reads. [OpenAlex](https://openalex.org/A5016877422) · [ORCID](https://orcid.org/0000-0002-3502-3574) · [**h-index 23**](https://openalex.org/authors/A5016877422) · [132 works](https://openalex.org/works?filter=author.id:A5016877422) · [2 357 citations](https://openalex.org/works?filter=author.id:A5016877422&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [6 works on this site](https://md.research.elamur.ai/free/en/authors/linfeng-song/#papers) · [869 citations in our library](https://md.research.elamur.ai/free/en/authors/linfeng-song/cited-by/) · 20 more works in our library · [⭐ Paper Star rank #2235](https://md.research.elamur.ai/free/en/paper-stars/page/23/#star-2235) 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=Linfeng+Song) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Linfeng+Song) · [dblp](https://dblp.org/search?q=Linfeng+Song) · [ACL Anthology](https://aclanthology.org/search/?q=Linfeng+Song) ## Affiliations - [Tencent (China)](https://md.research.elamur.ai/free/en/companies/tencent/) · Shenzhen, China —current · 2019–2026 on 32 papers - [University of Hong Kong](https://md.research.elamur.ai/free/en/universities/university-of-hong-kong/) · Pok Fu Lam, Hong Kong —current · 2026 on 1 paper - [Google (United States)](https://md.research.elamur.ai/free/en/companies/google/) · Mountain View, United States —2025 on 2 papers - [Kuaishou (China)](https://ror.org/0258as409) · Beijing —2023–2024 on 2 papers - [University of Rochester](https://md.research.elamur.ai/free/en/universities/university-of-rochester/) · Rochester, United States —2014–2021 on 17 papers · education 2014–2019: Ph.D., Computer Science (ORCID) - [Chinese Academy of Sciences](https://md.research.elamur.ai/free/en/government/chinese-academy-of-sciences/) · Beijing, China —2011–2021 on 6 papers - [City University of Hong Kong](https://md.research.elamur.ai/free/en/universities/city-university-of-hong-kong/) · Hong Kong, Hong Kong —2020–2021 on 4 papers - [Institute of Computing Technology](https://md.research.elamur.ai/free/en/labs/institute-of-computing-technology/) · Beijing, China —2011–2014 on 5 papers From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 20 more institutions on single bylines are not listed. ## Topics [Semantic parsing & meaning representation 3](https://md.research.elamur.ai/free/en/topics/semantic-parsing/) · [Machine translation 2](https://md.research.elamur.ai/free/en/topics/machine-translation/) · [Text generation 2](https://md.research.elamur.ai/free/en/topics/text-generation/) · [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/) · [Reasoning 1](https://md.research.elamur.ai/free/en/topics/reasoning/) ## Works by year ### 2020s [**2024** 1 paper](https://md.research.elamur.ai/free/en/papers/chen-2024-think-that-much-2-3/) · [**2022** 1 paper](https://md.research.elamur.ai/free/en/papers/xu-2022-guofeng-benchmark-zero-pronoun-recovery/) · [**2020** 1 paper](https://md.research.elamur.ai/free/en/papers/song-2020-zpr2-joint-zero-pronoun-recovery/) ### 2010s [**2019** 1 paper](https://md.research.elamur.ai/free/en/papers/song-2019-semantic-neural-machine-translation-amr/) · [**2018** 1 paper](https://md.research.elamur.ai/free/en/papers/song-2018-graph-sequence-model-amr-text/) · [**2017** 1 paper](https://md.research.elamur.ai/free/en/papers/song-2017-amr-text-generation-synchronous-node/) ## Papers The most cited first (OpenAlex and our library). 1. **[A Graph-to-Sequence Model for AMR-to-Text Generation](https://md.research.elamur.ai/free/en/papers/song-2018-graph-sequence-model-amr-text/)** *(2018 · 329 citations)* The paper addresses AMR-to-text generation, which means producing a sentence that expresses the meaning of an Abstract Meaning Representation graph. 2. **[Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs](https://md.research.elamur.ai/free/en/papers/chen-2024-think-that-much-2-3/)** *(2024 · 188 citations)* Reasoning models in the o1 family get better accuracy by writing very long chains of thought. This paper argues that much of that extra test-time compute is wasted. The authors call the waste "overthinking": after a model already has the right answer to a trivial question, it keeps producing new solution attempts.… 3. **[Semantic Neural Machine Translation Using AMR](https://md.research.elamur.ai/free/en/papers/song-2019-semantic-neural-machine-translation-amr/)** *(2019 · 159 citations)* The paper asks whether an AMR graph of the source sentence helps neural machine translation. The model keeps a BiLSTM encoder over the words and adds a graph recurrent network that encodes the AMR graph without linearising it. A doubly attentive LSTM decoder reads both encodings. On WMT16 English-German (4.17M… 4. **[AMR-to-text Generation with Synchronous Node Replacement Grammar](https://md.research.elamur.ai/free/en/papers/song-2017-amr-text-generation-synchronous-node/)** *(2017 · 108 citations)* The paper generates sentences directly from the AMR graph, without first making a spanning tree, so that no information is lost in the graph-to-tree step. 5. **[ZPR2: Joint Zero Pronoun Recovery and Resolution using Multi-Task Learning and BERT](https://md.research.elamur.ai/free/en/papers/song-2020-zpr2-joint-zero-pronoun-recovery/)** *(2020 · 29 citations)* Chinese often omits pronouns, especially in conversation, which hurts machine translation, QA and dialogue systems. 6. **[GuoFeng: A Benchmark for Zero Pronoun Recovery and Translation](https://md.research.elamur.ai/free/en/papers/xu-2022-guofeng-benchmark-zero-pronoun-recovery/)** *(2022 · 18 citations)* Pro-drop languages such as Chinese often leave pronouns unstated, and machine translation into English then produces output that is missing a subject or refers to the wrong entity. ## Also in our library 20 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. - [Yue Zhang](https://md.research.elamur.ai/free/en/authors/yue-zhang/) ×4 - [Daniel Gildea](https://md.research.elamur.ai/free/en/authors/daniel-gildea/) ×3 - [Zhiguo Wang](https://md.research.elamur.ai/free/en/authors/zhiguo-wang/) ×3 - [Zhaopeng Tu](https://md.research.elamur.ai/free/en/authors/zhaopeng-tu/) ×2 - [Derek F. Wong](https://md.research.elamur.ai/free/en/authors/derek-f-wong/) - [Dian Yu](https://md.research.elamur.ai/free/en/authors/dian-yu/) - [Dong Yu](https://md.research.elamur.ai/free/en/authors/dong-yu/) - [Jianshu Chen](https://md.research.elamur.ai/free/en/authors/jianshu-chen/) - [Jinsong Su](https://md.research.elamur.ai/free/en/authors/jinsong-su/) - [Kun Xu](https://md.research.elamur.ai/free/en/authors/kun-xu/) - [Lidia S. Chao](https://md.research.elamur.ai/free/en/authors/lidia-s-chao/) - [Longyue Wang](https://md.research.elamur.ai/free/en/authors/longyue-wang/) ## Cited by Papers of our library cite their works 869 times (of the 140 884 papers whose bibliographies we read); [67 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/linfeng-song/cited-by/).