Research map / Reinforcement learning / Imitation learning
Adversarial Soft Advantage Fitting: Imitation Learning without Policy Optimization
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 fixed-size sub-trajectories or single transitions instead of full trajectories (ablation) | ASAF · MuJoCo | performance | Helped |
| Use ASAF-1 or ASAF-w instead of full-trajectory ASAF (ablation) | ASAF · MuJoCo | learning performance | Helped |
| Use ASQF instead of ASAF-1 | transition-wise imitation learning · complex environments, including Pommerman | performance | Hurt |
| Add a gradient penalty to GAIL discriminator updates (ablation) | GAIL · MuJoCo | sample efficiency | Helped |
| Use ASAF, ASAF-w, or ASAF-1 rather than the baselines | imitation learning · Pommerman Random-Tag | evaluation return | Helped |
| Use ASAF-1 rather than the other evaluated methods | imitation learning · MuJoCo | training time | Helped |
| Use ASAF rather than ASAF-1 or ASAF-w (ablation) | ASAF · short classic-control and Box2D tasks | performance | No clear effect |
Closest papers by meaning
- Discriminator-Actor-Critic: Addressing Sample Inefficiency and Reward Bias in Adversarial Imitation Learning (ICLR 2019 · similarity 0.93)
- Deterministic and Discriminative Imitation (D2-Imitation): Revisiting Adversarial Imitation for Sample Efficiency (AAAI 2022 · similarity 0.92)
- Adversarial Imitation Learning via Boosting (ICLR 2024 · similarity 0.90)
- DPAIL: Training Diffusion Policy for Adversarial Imitation Learning without Policy Optimization (NeurIPS 2025 · similarity 0.90)
- Unlabeled Imperfect Demonstrations in Adversarial Imitation Learning (AAAI 2023 · similarity 0.90)
- Diffusion-Reward Adversarial Imitation Learning (NeurIPS 2024 · similarity 0.89)
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.