AAAI 2021 accepted papers on Large language models
77 of the 1,654 papers accepted at AAAI 2021 (4.7%) are on Large language models, compared with 6.6% at AAAI 2020. They fall into 4 topics.
See AAAI 2021 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Language, safety and interpretability · 44 papers
- UBAR: Towards Fully End-to-End Task-Oriented Dialog System with GPT-2 (123 citations)
- Learning Contextual Representations for Semantic Parsing with Generation-Augmented Pre-Training (76 citations)
- Re-TACRED: Addressing Shortcomings of the TACRED Dataset (71 citations)
- DialogBERT: Discourse-Aware Response Generation via Learning to Recover and Rank Utterances (69 citations)
- Automated Storytelling via Causal, Commonsense Plot Ordering (54 citations)
- FILTER: An Enhanced Fusion Method for Cross-lingual Language Understanding (54 citations)
- BERT & Family Eat Word Salad: Experiments with Text Understanding (51 citations)
- A Theoretical Analysis of the Repetition Problem in Text Generation (50 citations)
- Listen, Understand and Translate: Triple Supervision Decouples End-to-end Speech-to-text Translation (47 citations)
- Do Response Selection Models Really Know What’s Next? Utterance Manipulation Strategies for Multi-turn Response Selection (45 citations)
- Filling the Gap of Utterance-aware and Speaker-aware Representation for Multi-turn Dialogue (45 citations)
- XL-WSD: An Extra-Large and Cross-Lingual Evaluation Framework for Word Sense Disambiguation (45 citations)
and 32 more in the interactive map.
Retrieval and knowledge · 18 papers
- (Comet-) Atomic 2020: On Symbolic and Neural Commonsense Knowledge Graphs (293 citations)
- Retrospective Reader for Machine Reading Comprehension (191 citations)
- KG-BART: Knowledge Graph-Augmented BART for Generative Commonsense Reasoning (167 citations)
- UNICORN on RAINBOW: A Universal Commonsense Reasoning Model on a New Multitask Benchmark (78 citations)
- A Controllable Model of Grounded Response Generation (70 citations)
- Judgment Prediction via Injecting Legal Knowledge into Neural Networks (58 citations)
- EQG-RACE: Examination-Type Question Generation (42 citations)
- Knowledge-Driven Distractor Generation for Cloze-Style Multiple Choice Questions (37 citations)
- Learning to Copy Coherent Knowledge for Response Generation (26 citations)
- Reinforced History Backtracking for Conversational Question Answering (26 citations)
- Stratified Negation in Datalog with Metric Temporal Operators (16 citations)
- A Unified Pretraining Framework for Passage Ranking and Expansion (15 citations)
and 6 more in the interactive map.
AI and society · 11 papers
- Does Explainable Artificial Intelligence Improve Human Decision-Making? (106 citations)
- SCRUPLES: A Corpus of Community Ethical Judgments on 32,000 Real-Life Anecdotes (73 citations)
- Automated Lay Language Summarization of Biomedical Scientific Reviews (50 citations)
- Aligning Artificial Neural Networks and Ontologies towards Explainable AI (44 citations)
- A Novice-Reviewer Experiment to Address Scarcity of Qualified Reviewers in Large Conferences (23 citations)
- Argument Mining Driven Analysis of Peer-Reviews (17 citations)
- How Robust are Model Rankings : A Leaderboard Customization Approach for Equitable Evaluation (11 citations)
- We Can Explain Your Research in Layman’s Terms: Towards Automating Science Journalism at Scale (9 citations)
- Indecision Modeling (4 citations)
- AI-Assisted Scientific Data Collection with Iterative Human Feedback (2 citations)
- The Undergraduate Games Corpus: A Dataset for Machine Perception of Interactive Media (2 citations)
