The Evolution of Uncertainty of Learning in Games
Authors: Yun Kuen Cheung, Georgios Piliouras, Yixin Tao
ICLR 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | To illuminate this, we simulate MWU (with step-size 0.01) in Matching Pennies game, and present a few plots in Figure 2. The top-left plot shows the initial set, which is a small square around the unique Nash equilibrium. The top-right plot shows the range of possibility after 40000 steps. In our model, we assume the initial distribution is uniform over the small square, and plot the heat-map of probability densities after 40000 steps (bottom-left). We observe that the states in the boundary of the vortex are more probable to occur, while the densities around the Nash equilibrium decline. The bottom-right plot shows the densities that generate the heat-map. |
| Researcher Affiliation | Academia | Yun Kuen Cheung Royal Holloway University of London Georgios Piliouras Singapore University of Technology and Design Yixin Tao London School of Economics |
| Pseudocode | No | The paper provides mathematical equations for update rules (e.g., equations 1, 2, 3, 4, 5, 9, 10) but not structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any specific links to source code or make an explicit statement about releasing code for the methodology described. |
| Open Datasets | No | The paper describes theoretical analysis and simulations (e.g., 'simulate MWU (with step-size 0.01) in Matching Pennies game'). It assumes an 'initial distribution is uniform over the small square' for its simulation setup, which is a theoretical construct rather than a publicly available dataset. |
| Dataset Splits | No | No specific dataset split information (e.g., percentages, sample counts, or citations to predefined splits) for training, validation, or testing was provided, as the paper relies on theoretical analysis and simulations with an initial distribution, rather than traditional empirical data splits. |
| Hardware Specification | No | No specific hardware details (e.g., CPU/GPU models, memory, cloud instances) used for running the simulations or analysis were mentioned in the paper. |
| Software Dependencies | No | The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiments. |
| Experiment Setup | Yes | To illuminate this, we simulate MWU (with step-size 0.01) in Matching Pennies game, and present a few plots in Figure 2. The top-left plot shows the initial set, which is a small square around the unique Nash equilibrium. The top-right plot shows the range of possibility after 40000 steps. In our model, we assume the initial distribution is uniform over the small square, and plot the heat-map of probability densities after 40000 steps (bottom-left). We observe that the states in the boundary of the vortex are more probable to occur, while the densities around the Nash equilibrium decline. The bottom-right plot shows the densities that generate the heat-map. |