ICLR 2021 accepted papers on Large language models
15 of the 860 papers accepted at ICLR 2021 (1.7%) are on Large language models, compared with 3.3% at ICLR 2020. They fall into 4 topics.
See ICLR 2021 in the interactive map
Submitting to ICLR? Checklist →
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 · 9 papers
- Language-Agnostic Representation Learning of Source Code from Structure and Context (62 citations)
- FairFil: Contrastive Neural Debiasing Method for Pretrained Text Encoders (50 citations)
- SCoRe: Pre-Training for Context Representation in Conversational Semantic Parsing (44 citations)
- CoCo: Controllable Counterfactuals for Evaluating Dialogue State Trackers (41 citations)
- Pre-training Text-to-Text Transformers for Concept-centric Common Sense (35 citations)
- CoCon: A Self-Supervised Approach for Controlled Text Generation (16 citations)
- MIROSTAT: A NEURAL TEXT DECODING ALGORITHM THAT DIRECTLY CONTROLS PERPLEXITY (13 citations)
- Disambiguating Symbolic Expressions in Informal Documents
- Taking Notes on the Fly Helps Language Pre-Training
Agents, code and math · 4 papers
- Measuring Massive Multitask Language Understanding (325 citations)
- Mathematical Reasoning via Self-supervised Skip-tree Training (25 citations)
- Effective Abstract Reasoning with Dual-Contrast Network (9 citations)
- IsarStep: a Benchmark for High-level Mathematical Reasoning (8 citations)
AI and society · 1 paper
- Aligning AI With Shared Human Values (96 citations)
