--- title: "Karen Livescu — author profile and papers" description: "Karen Livescu, Toyota Technological Institute at Chicago: 5 papers on Elamur Research (2014–2021), h-index 47 — affiliations, topics, co-authors and who…" canonical: "https://research.elamur.ai/free/en/authors/karen-livescu/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/karen-livescu/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Karen Livescu # Karen Livescu Karen Livescu works on natural language processing, and the papers here focus on learned representations of words, phrases and sentences and on how those representations serve structured tasks. Recurring questions include whether continuous embeddings can stand in for older hand-built features, how noisy paraphrase resources can train general-purpose encoders, and how [coreference resolution](https://md.research.elamur.ai/free/en/topics/coreference-resolution/) systems can scale to long texts and carry over between corpora. The methods are mostly neural models trained and compared across several benchmark collections. The most cited of these papers is [Tailoring Continuous Word Representations for Dependency Parsing](https://md.research.elamur.ai/free/en/papers/bansal-2014-tailoring-continuous-word-representations-dependency/) (2014). It added embedding-based features to a second-order MSTParser and set several public and newly trained vector sets against Brown clusters for English [dependency parsing](https://md.research.elamur.ai/free/en/topics/dependency-parsing/). Two 2015 papers then used the Paraphrase Database as a source of supervision. [From Paraphrase Database to Compositional Paraphrase Model and Back](https://md.research.elamur.ai/free/en/papers/wieting-2015-paraphrase-database-compositional-paraphrase-model/) trained compact models, PARAGRAM word vectors and recursive networks, that can score any pair of phrases, and it introduced two crowd-annotated evaluation sets. [Towards Universal Paraphrastic Sentence Embeddings](https://md.research.elamur.ai/free/en/papers/wieting-2015-towards-universal-paraphrastic-sentence-embeddings/) compared six ways of composing word vectors into sentence encoders, from simple averaging to LSTMs, and tested them on in-domain data and 22 SemEval similarity sets. The later papers turn to coreference. [Learning to Ignore](https://md.research.elamur.ai/free/en/papers/toshniwal-2020-learning-ignore-long-document-coreference/) (2020) keeps only a fixed number of entity slots in memory, so runtime stays linear in the length of book-scale documents. [On Generalization in Coreference Resolution](https://md.research.elamur.ai/free/en/papers/toshniwal-2021-generalization-coreference-resolution/) (2021) brought eight datasets together and found that models trained on a single corpus transfer poorly. It reported gains from adding silver singleton mentions and from joint training on a sampled mix of corpora. Livescu's current institution is the [Toyota Technological Institute at Chicago](https://md.research.elamur.ai/free/en/universities/toyota-technological-institute-at-chicago/), with bylines there from 2008 onward. Earlier papers came from the [Massachusetts Institute of Technology](https://md.research.elamur.ai/free/en/universities/massachusetts-institute-of-technology/), including its Computer Science and Artificial Intelligence Laboratory. Frequent co-authors include [Kevin Gimpel](https://md.research.elamur.ai/free/en/authors/kevin-gimpel/), [Mohit Bansal](https://md.research.elamur.ai/free/en/authors/mohit-bansal/), John Wieting and Shubham Toshniwal. Karen Livescu has 5 papers in our library, in 2014–2021, most of them on [Coreference resolution](https://md.research.elamur.ai/free/en/topics/coreference-resolution/), [Dependency parsing](https://md.research.elamur.ai/free/en/topics/dependency-parsing/); often with [Kevin Gimpel](https://md.research.elamur.ai/free/en/authors/kevin-gimpel/), [Mohit Bansal](https://md.research.elamur.ai/free/en/authors/mohit-bansal/), [John Wieting](https://md.research.elamur.ai/free/en/authors/john-wieting/). 5 papers in our library · 2014–2021 current affiliation [Toyota Technological Institute at Chicago](https://md.research.elamur.ai/free/en/universities/toyota-technological-institute-at-chicago/), Chicago, United States [OpenAlex](https://openalex.org/A5015602781) · [ORCID](https://orcid.org/0000-0003-4962-946X) · [**h-index 47**](https://openalex.org/authors/A5015602781) · [257 works](https://openalex.org/works?filter=author.id:A5015602781) · [11 816 citations](https://openalex.org/works?filter=author.id:A5015602781&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [5 works in our library](https://md.research.elamur.ai/free/en/authors/karen-livescu/#papers) · [1 804 citations in our library](https://md.research.elamur.ai/free/en/authors/karen-livescu/cited-by/) · 5 more works in our library · [⭐ Paper Star rank #1749](https://md.research.elamur.ai/free/en/paper-stars/page/18/#star-1749) 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=Karen+Livescu) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Karen+Livescu) · [dblp](https://dblp.org/search?q=Karen+Livescu) · [ACL Anthology](https://aclanthology.org/search/?q=Karen+Livescu) ## Affiliations - [Toyota Technological Institute at Chicago](https://md.research.elamur.ai/free/en/universities/toyota-technological-institute-at-chicago/) · Chicago, United States —current · 2008–2026 on 144 papers - [University of Chicago](https://md.research.elamur.ai/free/en/universities/university-of-chicago/) · Chicago, United States —2022–2024 on 4 papers - [Carnegie Mellon University](https://md.research.elamur.ai/free/en/universities/carnegie-mellon-university/) · Pittsburgh, United States —2024 on 3 papers - [Massachusetts Institute of Technology](https://md.research.elamur.ai/free/en/universities/massachusetts-institute-of-technology/) · Cambridge, United States —2001–2012 on 18 papers - [MIT Computer Science and Artificial Intelligence Laboratory](https://md.research.elamur.ai/free/en/labs/mit-computer-science-and-artificial-intelligence-laboratory/) · Cambridge, United States —2004–2007 on 7 papers From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 6 more institutions on single bylines are not listed. ## Topics [Coreference resolution 2](https://md.research.elamur.ai/free/en/topics/coreference-resolution/) · [Dependency parsing 1](https://md.research.elamur.ai/free/en/topics/dependency-parsing/) ## Works by year ### 2020s [**2021** 1 paper](https://md.research.elamur.ai/free/en/papers/toshniwal-2021-generalization-coreference-resolution/) · [**2020** 1 paper](https://md.research.elamur.ai/free/en/papers/toshniwal-2020-learning-ignore-long-document-coreference/) ### 2010s [**2015** 2 papers](https://md.research.elamur.ai/free/en/authors/karen-livescu/year/2015/) · [**2014** 1 paper](https://md.research.elamur.ai/free/en/papers/bansal-2014-tailoring-continuous-word-representations-dependency/) ## Papers The most cited first (OpenAlex and our library). 1. **[Tailoring Continuous Word Representations for Dependency Parsing](https://md.research.elamur.ai/free/en/papers/bansal-2014-tailoring-continuous-word-representations-dependency/)** *(2014 · 418 citations)* The paper asks whether continuous word embeddings can give English dependency parsing the same kind of boost that Brown clusters already provide. 2. **[From Paraphrase Database to Compositional Paraphrase Model and Back](https://md.research.elamur.ai/free/en/papers/wieting-2015-paraphrase-database-compositional-paraphrase-model/)** *(2015 · 375 citations)* PPDB is a large, automatically built collection of paraphrase pairs, but it has three weaknesses. Two phrases can only be compared if both appear in it. Its size grows with the source data. Its confidence scores come from hand-weighted features of uncertain quality. Wieting et al. use PPDB as training data for… 3. **[Towards Universal Paraphrastic Sentence Embeddings](https://md.research.elamur.ai/free/en/papers/wieting-2015-towards-universal-paraphrastic-sentence-embeddings/)** *(2015 · 272 citations)* Word embeddings transfer easily across tasks, but comparably reusable sentence vectors were still lacking. 4. **[Learning to Ignore: Long Document Coreference with Bounded Memory Neural Networks](https://md.research.elamur.ai/free/en/papers/toshniwal-2020-learning-ignore-long-document-coreference/)** *(2020 · 80 citations)* Leading coreference systems need memory and runtime that grow quadratically with document length, which makes book-scale texts impractical. 5. **[On Generalization in Coreference Resolution](https://md.research.elamur.ai/free/en/papers/toshniwal-2021-generalization-coreference-resolution/)** *(2021 · 62 citations)* The paper asks whether coreference models trained on one dataset transfer to others. It consolidates eight datasets (OntoNotes, LitBank, PreCo as training sets; Character Identification, WikiCoref, Quiz Bowl, GAP and Winograd-style WSC as evaluation or analysis sets) and trains a long-document entity-ranking model… ## 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 7 co-authors with a page, the most papers together first. - [Kevin Gimpel](https://md.research.elamur.ai/free/en/authors/kevin-gimpel/) ×5 - [Mohit Bansal](https://md.research.elamur.ai/free/en/authors/mohit-bansal/) ×3 - [John Wieting](https://md.research.elamur.ai/free/en/authors/john-wieting/) ×2 - [Sam Wiseman](https://md.research.elamur.ai/free/en/authors/sam-wiseman/) ×2 - [Shubham Toshniwal](https://md.research.elamur.ai/free/en/authors/shubham-toshniwal/) ×2 - [Allyson Ettinger](https://md.research.elamur.ai/free/en/authors/allyson-ettinger/) - [Patrick Xia](https://md.research.elamur.ai/free/en/authors/patrick-xia/) ## Cited by Papers of our library cite their works 1 804 times (of the 140 884 papers whose bibliographies we read); [205 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/karen-livescu/cited-by/).