Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Inverse Constrained Reinforcement Learning
Authors: Shehryar Malik, Usman Anwar, Alireza Aghasi, Ali Ahmed
ICML 2021 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We experimentally validate our approach and show that our framework can successfully learn the most likely constraints that the agent respects. |
| Researcher Affiliation | Academia | 1Information Technology University, Lahore, Pakistan 2Georgia State University, Atlanta, GA, USA. |
| Pseudocode | Yes | Algorithm 1 ICRL |
| Open Source Code | Yes | The code is available at: https: //github.com/shehryar-malik/icrl. |
| Open Datasets | Yes | For this experiment, we use two simulated robots called Half Cheetah and Ant from Open AI Gym (Brockman et al., 2016). |
| Dataset Splits | No | The paper does not explicitly provide training/validation/test dataset splits. It describes using environments from OpenAI Gym, which typically involve continuous interaction rather than predefined static datasets with explicit splits. |
| Hardware Specification | No | The paper does not provide specific details about the hardware used for running the experiments (e.g., CPU, GPU models, or memory). |
| Software Dependencies | No | The paper mentions using Stable Baselines and Weights and Biases, but it does not specify version numbers for these or other core software components like Python, PyTorch/TensorFlow, or CUDA. |
| Experiment Setup | Yes | Finally, all hyperparameters and additional details of the experiments can be found in Section 8.5 in the supplementary material. |