Research map / Reinforcement learning / Imitation learning
Provable Guarantees for Generative Behavior Cloning: Bridging Low-Level Stability and High-Level Behavior
Adam Block, Ali Jadbabaie, Daniel Pfrommer, Max Simchowitz, Russ Tedrake
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Diffuse stabilizing gain matrices instead of raw control inputs | DDPM behavior cloning on the 2-D quadrotor · 2-D quadrotor | performance | Improved |
| Use one demonstration with synthesized gains and data smoothing instead of ten demonstrations without intervention | DDPM behavior cloning on the 2-D quadrotor · 2-D quadrotor | performance | Improved |
| Add observation-smoothing noise back at inference time | Hint · PushT | test performance | Improved |
| Add observation-smoothing noise back at inference time | Hint · Can Pick-and-Place | test performance | Improved |
| Add observation-smoothing noise back at inference time | Hint · Square Nut Assembly | test performance | Improved |
| Use the smoothed replica policy rather than the unnoised policy | Hint · PushT | test performance | Improved |
| Use the smoothed replica policy rather than the unnoised policy | Hint · Can Pick-and-Place, low-data regime | test performance | Improved |
| Use the smoothed replica policy rather than the unnoised policy | Hint · Robotic manipulation tasks | test performance | No clear effect |
| Apply data smoothing without diffusing stabilizing gain matrices (ablation) | Raw-input DDPM behavior cloning · 2-D quadrotor | performance | Hurt |
Closest papers by meaning
- Diffusion Model-Augmented Behavioral Cloning (ICML 2024 · similarity 0.91)
- Learning When to Stop: Selective Imitation Learning Under Arbitrary Dynamics Shift (NeurIPS 2026 · similarity 0.90)
- Is Behavior Cloning All You Need? Understanding Horizon in Imitation Learning (NeurIPS 2024 · similarity 0.90)
- Interactive and Hybrid Imitation Learning: Provably Beating Behavior Cloning (NeurIPS 2025 · similarity 0.90)
- Improving Generative Behavior Cloning via Self-Guidance and Adaptive Chunking (NeurIPS 2025 · similarity 0.90)
- CCIL: Continuity-Based Data Augmentation for Corrective Imitation Learning (ICLR 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.
