DeeLIO @ NAACL 2021

7 papers

Top Research Directions

The 25 research directions with DeeLIO @ NAACL 2021 papers, ranked by paper count. Share is the fraction of the edition’s papers in the direction.

#DirectionPapersShare
1Hierarchical Feature Abstraction Across Transformer Layers228.6%
2Multi-Hop Reasoning over Knowledge Graphs228.6%
3The Incompleteness Bottleneck of External Knowledge Graphs228.6%
4Bridging the Commonsense Reasoning Gap in Language Models114.3%
5Bridging the Semantic Gap in Natural Language Interfaces114.3%
6Concept Based Explanations for Deep Learning Models114.3%
7Decomposing Context into Multi-Level Attention Hierarchies114.3%
8Decomposing High Dimensional Embeddings into Interpretable Concepts114.3%
9Denoising Subgraphs for Knowledge Augmented Reasoning114.3%
10Distributed Semantic Mapping Across the Human Cortex114.3%
11Domain Adaptation under Extreme Target Data Scarcity114.3%
12Dynamic Graph Evolution for Conversational Modeling114.3%
13Emergent Functional Specialization in Deep Neural Networks114.3%
14Enriching Sparse Text Representations with External Knowledge114.3%
15Evaluating Logical Validity Beyond Surface Text Overlap114.3%
16General Purpose Reasoning from Specialized Reinforcement Learning114.3%
17Graph Based Dialogue Representation114.3%
18Grounding Language Models with Structured Knowledge Graphs114.3%
19Iterative Retrieval Augmented Generation for Multi Hop Reasoning114.3%
20Modeling Neighborhood Context in Knowledge Graphs114.3%
21Reasoning Over Incomplete Knowledge Graphs114.3%
22Reconstructing Implicit Premises in Natural Language Arguments114.3%
23Reducing the Computational Cost of Post-Hoc Feature Decomposition114.3%
24Serializing Structured Graphs into Natural Language Sequences114.3%
25Unsupervised Concept Discovery in Deep Learning114.3%

Download all directions as CSV

Institutions

Institutions credited by DeeLIO @ NAACL 2021 papers: a paper credits every institution named in its affiliations, once. 15 institutions share 7 credited papers.

#InstitutionPapersShare
1Carnegie Mellon University 228.6%
2Google (United States) 114.3%
3Heidelberg University 114.3%
4Hong Kong University of Science and Technology 114.3%
5Institute for Computational Linguistics "A. Zampolli" 114.3%
6Lawrence Berkeley National Laboratory 114.3%
7Meta (United States) 114.3%
8New York University 114.3%
9Scuola Normale Superiore 114.3%
10Tsinghua University 114.3%
11University of British Columbia 114.3%
12University of California, Berkeley 114.3%
13University of North Carolina at Chapel Hill 114.3%
14University of Pisa 114.3%
15University of Pittsburgh 114.3%

Download all institutions as CSV

About this data

Papers. DeeLIO @ NAACL 2021 lists 7 papers on Lacuna, duplicate versions counted once; rejected and withdrawn submissions are left out.

Directions. Directions are Lacuna’s clustering of the concept elements extracted from each paper, so a paper can belong to several and direction counts add up to more than the paper count. A direction’s share is its papers divided by the edition’s papers.

Institutions. A paper credits each institution named in its affiliations, resolved to a ROR record, once. 7 of 7 papers (100.0%) credit at least one institution and are the base for institution shares; the rest have no resolved affiliation yet.

Data. Counts come from Lacuna’s data build of 2026-11-19 and change as papers are added. Tables as CSV: directions · institutions. Cite as: Lacuna, “DeeLIO @ NAACL 2021”, https://lacuna.tiptreesystems.com/venue/deelio-naacl/2021.

Top Authors

AuthorPapers
Alessio Miaschi1
Bhuwan Dhingra1
Bruno Olshausen1
Giovanni Puccetti1
Hyeju Jang1
Mohit Bansal1
Raymond T. Ng1
Vidhisha Balachandran1
Wen Xiao1
William W. Cohen1
Yann LeCun1
Yubei Chen1
Zeyu Yun1

Impactful Papers

Papers from DeeLIO @ NAACL 2021 that shaped research clusters.

Foundational

Influential

Papers

Title
Augmenting Topic Aware Knowledge-Grounded Conversations with Dynamic Built Knowledge Graphs
ERNIE-NLI: Analyzing the Impact of Domain-Specific External Knowledge on Enhanced Representations for NLI
How Do BERT Embeddings Organize Linguistic Knowledge?
Investigating the Effect of Background Knowledge on Natural Questions
KW-ATTN: Knowledge Infused Attention for Accurate and Interpretable Text Classification
Reconstructing Implicit Knowledge with Language Models
Transformer visualization via dictionary learning: contextualized embedding as a linear superposition of transformer factors