Research map / Reinforcement learning / Policy gradient and actor-critic
TraCeS: Learning Per-Timestep Constraint-Violation Credit from Sparse Trajectory-Level Labels
Siow Meng Low, Ze Gong, Akshat Kumar
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 |
|---|---|---|---|
| use TraCeS per-timestep credit instead of the C-T MLP estimator | PPO-Lagrangian · PointCircle1 | labeled trajectories: 12000 -> 5830 (918) | Improved |
| use TraCeS instead of C-T | PPO-Lagrangian · Hopper | cost: 284.2 (289.2) -> 17.1 (4.7) | Improved |
| use TraCeS instead of C-T | PPO-Lagrangian · HalfCheetah | cost: 99.7 (74.5) -> 22.4 (5.3) | Improved |
| use TraCeS instead of C-T | PPO-Lagrangian · CarCircle2 | cost: 44.0 (24.0) -> 11.9 (10.0) | Improved |
| use TraCeS instead of C-T | PPO-Lagrangian · Ant | labeled trajectories: 6000 (0) -> 4657 (551) | Improved |
| use TraCeS instead of C-T | PPO-Lagrangian · Walker2d | cost: 16.7 (9.1) -> 24.1 (3.9) | Hurt |
| use TraCeS instead of partial-knowledge RLSF | PPO-Lagrangian · Walker2d | labeled trajectories: 15227 (1920) -> 5735 (1041) | Improved |
| use TraCeS instead of partial-knowledge RLSF | PPO-Lagrangian · PointCircle1 | cost: 1.6 (2.2) -> 16.2 (7.2) | Hurt |
| use TraCeS instead of RLSF | PPO-Lagrangian · CarCircle1 | cost: 33.8 (41.7) -> 11.3 (8.0) | Improved |
Closest papers by meaning
- Imitate the Good and Avoid the Bad: An Incremental Approach to Safe Reinforcement Learning (AAAI 2024 · similarity 0.91)
- Near-Optimal Sample Complexity for Online Constrained MDPs (NeurIPS 2025 · similarity 0.89)
- Reward Penalties on Augmented States for Solving Richly Constrained RL Effectively (AAAI 2024 · similarity 0.89)
- Offline Safe Reinforcement Learning Using Trajectory Classification (AAAI 2025 · similarity 0.88)
- Safety Representations for Safer Policy Learning (ICLR 2025 · similarity 0.88)
- Safe Reinforcement Learning with Instantaneous Constraints: The Role of Aggressive Exploration (AAAI 2024 · similarity 0.88)
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.
