Research map / Reinforcement learning / Imitation learning
State-only Imitation with Transition Dynamics Mismatch
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 I2L instead of GAIL-S with matched transition dynamics | State-only imitation on MuJoCo locomotion tasks · Walker2d, Hopper, Ant, and Half-Cheetah (matched dynamics) | average episodic return: 3711 -> 4107; 2130 -> 2751; 3217 -> 3320; 5974 -> 5240 | Mixed |
| Use I2L instead of GAIfO | State-only imitation on MuJoCo locomotion tasks · Matched dynamics, low gravity, high density, and high friction | average episodic return: 5082 -> 5240; 3122 -> 4107; 2121 -> 2751; 3452 -> 3320; 1518 -> 4155; 2995 -> 3547; 1683 -> 2566; 594 -> 1617; -234 -> 3 | Helped |
| Use I2L instead of GAIL-S under transition dynamics mismatch | State-only imitation on MuJoCo locomotion tasks · Mismatched dynamics, including low gravity, high density, and high friction | average episodic return: 923 -> 3975; 810 -> 2084 | Helped |
| Increase the I2L trajectory-buffer capacity to 50 (ablation) | I2L · Half-Cheetah (low gravity) | average episodic return | No clear effect |
| Reduce the I2L trajectory-buffer capacity to 1 (ablation) | I2L · Half-Cheetah (low gravity) | average episodic return | No clear effect |
Closest papers by meaning
- Interactive and Hybrid Imitation Learning: Provably Beating Behavior Cloning (NeurIPS 2025 · similarity 0.91)
- Markov Balance Satisfaction Improves Performance in Strictly Batch Offline Imitation Learning (AAAI 2025 · similarity 0.90)
- Diffusion Imitation from Observation (NeurIPS 2024 · similarity 0.90)
- Is Behavior Cloning All You Need? Understanding Horizon in Imitation Learning (NeurIPS 2024 · similarity 0.90)
- Causal Imitation for Markov Decision Processes: a Partial Identification Approach (NeurIPS 2024 · similarity 0.90)
- Diffusion Model-Augmented Behavioral Cloning (ICML 2024 · similarity 0.90)
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.