Research map / Adversarial robustness and security / Attacks on text and graphs
End-to-End Differentiable Adversarial Imitation Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use MGAIL instead of GAIL | MGAIL · Ant, 4 trajectories | total reward: 3186.80 ± 903.57 -> 4645.12 ± 179.29 | Helped |
| use MGAIL instead of GAIL | MGAIL · Walker, 4 trajectories | total reward: 4877.98 ± 2848.37 -> 6916.34 ± 115.20 | Helped |
| use MGAIL instead of GAIL | MGAIL · Half-Cheetah, 4 trajectories | total reward: 4515.70 ± 549.49 -> 4891.56 ± 654.43 | Helped |
| use MGAIL instead of GAIL | MGAIL · Humanoid, 240 trajectories | total reward: 10361.94 ± 61.28 -> 10470.94 ± 54.35 | Helped |
| use MGAIL instead of GAIL | MGAIL · Cartpole, 1 trajectory | total reward: 200.00 ± 0.00 -> 200.00 ± 0.00 | No clear effect |
| use MGAIL instead of GAIL | MGAIL · Mountain Car, 1 trajectory | total reward: -101.55 ± 10.32 -> -107.4 ± 10.89 | Hurt |
| use MGAIL instead of GAIL | MGAIL · Acrobot, 1 trajectory | total reward: -77.26 ± 18.03 -> -85.65 ± 23.74 | Hurt |
| use MGAIL instead of GAIL | MGAIL · Mountain Car, 7 trajectories | total reward: -99.90 ± 7.97 -> -104.23 ± 14.31 | Hurt |
| use adversarial imitation (MGAIL) instead of behavioral cloning | MGAIL · Hopper, 4 trajectories | total reward: 50.57 ± 0.95 -> 3669.53 ± 6.09 | Helped |
| use adversarial imitation (MGAIL) instead of behavioral cloning | MGAIL · Walker, 25 trajectories | total reward: 1599.36 ± 1456.59 -> 7070.45 ± 30.68 | Helped |
| increase expert trajectories for behavioral cloning from 4 to 25 | Behavioral cloning · Hopper | total reward: 50.57 ± 0.95 -> 3383.96 ± 657.61 | Helped |
| use advanced forward model instead of basic forward model (ablation) | MGAIL · MuJoCo tasks | total reward | Helped |
Closest papers by meaning
- Imitation Learning via Kernel Mean Embedding (AAAI 2018 · similarity 0.89)
- Generative Adversarial Imitation Learning (NeurIPS 2016 · similarity 0.89)
- A Bayesian Approach to Generative Adversarial Imitation Learning (NeurIPS 2018 · similarity 0.88)
- Diffusing States and Matching Scores: A New Framework for Imitation Learning (ICLR 2025 · similarity 0.88)
- Exploring Gradient Explosion in Generative Adversarial Imitation Learning: A Probabilistic Perspective (AAAI 2024 · similarity 0.87)
- Diffusion Imitation from Observation (NeurIPS 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.