ICLR 2020 accepted papers by research area
687 papers were accepted at ICLR 2020 (+185 vs 2019). Here is how they split across 20 research areas, and which topics gained the most ground.
See ICLR 2020 in the interactive map
Submitting to ICLR? Checklist →
Research areas
| Area | Papers | Share | Change in share vs 2019 |
|---|---|---|---|
| Architectures and efficiency | 121 | 17.6% | +1.1 pts |
| Reinforcement learning | 81 | 11.8% | -0.8 pts |
| Label-efficient and robust learning | 79 | 11.5% | +0.3 pts |
| Probabilistic and scientific modeling | 59 | 8.6% | -3.6 pts |
| Learning theory and algorithms | 54 | 7.9% | +0.5 pts |
| Optimization | 52 | 7.6% | -1.0 pts |
| Graph learning | 49 | 7.1% | +3.5 pts |
| Adversarial robustness and security | 49 | 7.1% | +1.4 pts |
| Multi-agent systems and games | 29 | 4.2% | +1.2 pts |
| Image generation | 23 | 3.3% | -4.2 pts |
| Large language models | 23 | 3.3% | +1.2 pts |
| Detection and segmentation | 14 | 2.0% | -0.2 pts |
| Video | 12 | 1.7% | +0.6 pts |
| 3D vision | 11 | 1.6% | +0.2 pts |
| Continual learning | 10 | 1.5% | +0.3 pts |
| Federated learning and privacy | 6 | 0.9% | +0.5 pts |
| Bandits and online learning | 5 | 0.7% | +0.3 pts |
| Vision-language and multimodal | 4 | 0.6% | -1.0 pts |
| Point clouds | 3 | 0.4% | steady |
| Time series | 3 | 0.4% | -0.4 pts |
Fastest-growing topics at ICLR 2020
Change in share of ICLR papers from 2019 to 2020, topics with at least 10 papers.
- GNN architectures (Graph learning): 7 → 22 papers
- Self-supervised and contrastive (Label-efficient and robust learning): 0 → 10 papers
- Attention and transformers (Architectures and efficiency): 8 → 20 papers
- CNN design (Architectures and efficiency): 11 → 23 papers
- Attacks on text and graphs (Adversarial robustness and security): 4 → 11 papers
- Language, safety and interpretability (Large language models): 7 → 15 papers
- Human-AI and goal-conditioned (Multi-agent systems and games): 10 → 19 papers
- Knowledge and temporal graphs (Graph learning): 5 → 12 papers
- Neural architecture search (Architectures and efficiency): 6 → 13 papers
- Policy gradient and actor-critic (Reinforcement learning): 13 → 22 papers
More venues
Same year: ICML 2020 · NeurIPS 2020 · CVPR 2020 · AAAI 2020 · ECCV 2020 · TPAMI 2020 · IJCV 2020 · JMLR 2020
ICLR in other years: 2017 · 2018 · 2019 · 2021 · 2022 · 2023 · 2024 · 2025 · 2026
