Opponent Modeling based on Subgoal Inference

Authors: XiaoPeng Yu, Jiechuan Jiang, Zongqing Lu

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirically, we show that our method achieves more effective adaptation than existing methods in a variety of tasks.
Researcher Affiliation Academia Xiaopeng Yu Jiechuan Jiang Zongqing Lu School of Computer Science, Peking University
Pseudocode Yes For completeness, the full procedure of OMG is given in Algorithm 1.
Open Source Code Yes To ensure reproducibility, we include the code in the supplementary material and will make it open-source upon acceptance.
Open Datasets Yes Foraging [2, 4] is an 8 8 gridworld... Predator-Prey [20] is a three-against-one multi-agent environment... SMAC [35] is a high-dimensional environment for collaborative multi-agent reinforcement learning based on Star Craft II
Dataset Splits No The paper mentions 'training set' and 'test set' but does not explicitly specify a separate 'validation' set or its split percentages/counts for model selection during training.
Hardware Specification Yes The computational resources for the experiments are as follows: the CPU is Intel(R) Xeon(R) Platinum 8280 CPU @ 2.70GHz, and the GPU is A100-PCIE-40GB.
Software Dependencies No Table 1 lists various components like 'MLP', 'RNN', 'Re LU', 'Adam', 'RMSProp', 'CVAE' but does not provide specific version numbers for software libraries or dependencies (e.g., PyTorch 1.x, Python 3.x).
Experiment Setup Yes All hyperparameters are listed in Table 1.