AAAI 2021 accepted papers by research area
1,654 papers were accepted at AAAI 2021 (+44 vs 2020). Here is how they split across 20 research areas, and which topics gained the most ground.
See AAAI 2021 in the interactive map
Research areas
| Area | Papers | Share | Change in share vs 2020 |
|---|---|---|---|
| Label-efficient and robust learning | 262 | 15.8% | -1.3 pts |
| Learning theory and algorithms | 234 | 14.1% | +0.1 pts |
| Architectures and efficiency | 128 | 7.7% | +0.3 pts |
| Graph learning | 126 | 7.6% | -0.9 pts |
| Probabilistic and scientific modeling | 119 | 7.2% | +1.5 pts |
| Detection and segmentation | 88 | 5.3% | -0.7 pts |
| Reinforcement learning | 86 | 5.2% | +0.9 pts |
| Large language models | 77 | 4.7% | -2.0 pts |
| Multi-agent systems and games | 77 | 4.7% | +0.1 pts |
| Vision-language and multimodal | 67 | 4.1% | +0.3 pts |
| Image generation | 64 | 3.9% | -1.5 pts |
| Adversarial robustness and security | 56 | 3.4% | +0.3 pts |
| Video | 55 | 3.3% | -1.3 pts |
| Bandits and online learning | 43 | 2.6% | +1.0 pts |
| Optimization | 42 | 2.5% | +0.9 pts |
| Time series | 32 | 1.9% | +0.8 pts |
| 3D vision | 31 | 1.9% | -0.4 pts |
| Federated learning and privacy | 27 | 1.6% | +0.7 pts |
| Continual learning | 20 | 1.2% | +0.7 pts |
| Point clouds | 20 | 1.2% | +0.5 pts |
Fastest-growing topics at AAAI 2021
Change in share of AAAI papers from 2020 to 2021, topics with at least 10 papers.
- Molecules and proteins (Probabilistic and scientific modeling): 12 → 46 papers
- Prediction and decision losses (Learning theory and algorithms): 25 → 52 papers
- Visual and representation RL (Reinforcement learning): 22 → 42 papers
- Human and embodied modeling (Vision-language and multimodal): 20 → 38 papers
- Self-supervised and contrastive (Label-efficient and robust learning): 4 → 19 papers
- Optimizers (Optimization): 10 → 22 papers
- Policy gradient and actor-critic (Reinforcement learning): 17 → 26 papers
- Certified and empirical robustness (Adversarial robustness and security): 16 → 24 papers
- Human-AI and goal-conditioned (Multi-agent systems and games): 29 → 37 papers
- Contextual bandits and Thompson sampling (Bandits and online learning): 6 → 13 papers
More venues
Same year: ICML 2021 · NeurIPS 2021 · ICLR 2021 · CVPR 2021 · ICCV 2021 · TPAMI 2021 · IJCV 2021 · JMLR 2021
AAAI in other years: 2016 · 2017 · 2018 · 2019 · 2020 · 2022 · 2023 · 2024 · 2025 · 2026
