Research map / Reinforcement learning / Imitation learning
Adversarial imitation learning
Match expert behavior with GAN-style discriminators and provably efficient adversarial objectives.
33 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Imitation learning, keyword cluster “il · adversarial imitation · cloning”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
8 papers in 2023–24 and 8 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic held steady, from 6.7% to 6.9%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Generative Adversarial Imitation Learning (NeurIPS 2016 · 353 citations)
- 2017 · Robust Imitation of Diverse Behaviors (NeurIPS 2017 · 114 citations)
- 2018 · A Bayesian Approach to Generative Adversarial Imitation Learning (NeurIPS 2018 · 15 citations)
- 2019 · Variational Discriminator Bottleneck: Improving Imitation Learning, Inverse RL, and GANs by Constraining Information Flow (ICLR 2019 · 143 citations)
- 2020 · f-GAIL: Learning f-Divergence for Generative Adversarial Imitation Learning (NeurIPS 2020 · 12 citations)
- 2021 · What Matters for Adversarial Imitation Learning? (NeurIPS 2021 · 18 citations)
- 2022 · Deterministic and Discriminative Imitation (D2-Imitation): Revisiting Adversarial Imitation for Sample Efficiency (AAAI 2022 · 8 citations)
- 2023 · Unlabeled Imperfect Demonstrations in Adversarial Imitation Learning (AAAI 2023 · 9 citations)
- 2024 · Exploring Gradient Explosion in Generative Adversarial Imitation Learning: A Probabilistic Perspective (AAAI 2024 · 12 citations)
- 2025 · POI Recommendation via Multi-Objective Adversarial Imitation Learning (AAAI 2025 · 2 citations)
- 2026 · Non-Adversarial Imitation Learning Provably Free of Compounding Errors: The Value Flow Mechanism (ICML 2026 · 0 citations)
Earliest papers here
- Generative Adversarial Imitation Learning (NeurIPS 2016 · 353 citations)
- Robust Imitation of Diverse Behaviors (NeurIPS 2017 · 114 citations)
- A Bayesian Approach to Generative Adversarial Imitation Learning (NeurIPS 2018 · 15 citations)
What has been tried
24 papers here report 212 measured changes (a component added, removed or swapped, compared with a baseline): helped 64%, no clear effect 9%, hurt 18%, mixed 9%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Increase diffusion discriminator steps from 2 to 10 | MuJoCo tasks · training stability and convergence speed | Helped | DiffAIL: Diffusion Adversarial Imitation Learning (AAAI 2024 · 10 citations) | |
| Vary controller strength k among 0.1, 1, and 10 | MuJoCo tasks · normalized return | Mixed | C-GAIL: Stabilizing Generative Adversarial Imitation Learning with Control Theory (NeurIPS 2024 · 3 citations) | |
| Add multi-reward learning | Tokyo · Acc@5 | 17% performance drop (Acc@5 in TKY) when removed | Helped | POI Recommendation via Multi-Objective Adversarial Imitation Learning (AAAI 2025 · 2 citations) |
| increase policy updates per discriminator update | Walker Walk and Ball in Cup Catch · sample efficiency | Hurt | Adversarial Imitation Learning via Boosting (ICLR 2024 · 0 citations) |
What reviewers will expect
Datasets and baselines named in the 7 papers of this approach from 2024–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Baselines
- GAIL · 3 of 7 papers
- Behavior cloning · 2 of 7 papers
- ValueDICE · 2 of 7 papers
- AIRL · 2 of 7 papers
- WAIL · 2 of 7 papers
- DiffAIL · 2 of 7 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 16 papers; from OpenAlex and the papers' affiliations).
- Tsinghua University · 2 papers
Use it
Download BibTeX of 28 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Offline imitation from imperfect data (108 papers)
- Imitation combined with RL (108 papers)
- Behavior cloning theory and models (68 papers)
Elsewhere in the topic
- Reward learning from demonstrations (70 papers)
- Theory of inverse RL (44 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).