Udrea, Dorin-Vlad (2025) CABO - a Computational Analysis of Belief and Opponent-modeling. Bachelor's Thesis, Artificial Intelligence.
|
Text
s4645014.pdf Download (480kB) | Preview |
|
![[img]](https://fse.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
Toestemming.pdf Restricted to Registered users only Download (190kB) |
Abstract
The human capacity to reason about the unobservable mental states of others, known as Theory of Mind (ToM), is fundamental to navigating complex social interactions. This thesis investigates the strategic value of ToM in a noisy, incomplete-information environment by designing and evaluating a hierarchy of computational agents for the card game CABO. We created a series of AI players with increasingly advanced strategic thinking, beginning with a simple agent that only reacted to observable game events and progressing to complex agents that could model their opponents’ hidden beliefs and even anticipate deceptive plays. Agent performance was evaluated in both direct one-on-one competition and complex multi-agent free-for-all games. The results demonstrate a clear, context-dependent performance hierarchy. In pairwise simulations, higher-order ToM confers a significant strategic advantage, with more sophisticated agents consistently outperforming simpler ones. However, this advantage diminishes and even inverts in multi-agent scenarios, where the increased social complexity and cognitive load appear to favor more robust, less complex strategies. The findings confirm that the strategic value of ToM is not absolute but is contingent on the social context and the cognitive capabilities of one’s opponents, providing a computational framework for exploring the nuanced dynamics of social intelligence.
| Item Type: | Thesis (Bachelor's Thesis) |
|---|---|
| Supervisor name: | Weerd, H.A. de |
| Degree programme: | Artificial Intelligence |
| Thesis type: | Bachelor's Thesis |
| Language: | English |
| Date Deposited: | 09 Sep 2025 10:02 |
| Last Modified: | 23 Sep 2025 07:46 |
| URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/36875 |
Actions (login required)
 |
View Item |