Large language models enable rich communication and flexible planning in embodied agents, yet their updates to internal state remain correlation driven, making them prone to hallucinations that undermine reliability in interactive settings. To address this limitation, we investigate whether a true causal model can replace template based belief, desire, and intention (BDI) updates and provide more stable reasoning. We introduce CausalMind, an embodied agent architecture that integrates a causal model trained on text based BDI transitions drawn from a synthetic dataset. We evaluate both the causal model and the full agent in iTHOR across solo and cooperative tasks.

The results indicate that learning causal variables from text embeddings is difficult. The model focused primarily on action related features, failing to separate belief, desire, and intention and instead overfitting to biases in the dataset. Expanding and refining the synthetic dataset improved performance but did not resolve the core issues, highlighting the need for richer data and more suitable encoders. At the agent level, perception errors from the vision module frequently led to hallucinated task completion, limiting the usefulness of teacher guidance and making complex tasks difficult in the iTHOR environment.

Although the current implementation does not outperform template based baselines, the study clarifies the obstacles that arise when extending causal representation learning to natural language BDI states. These findings outline the key requirements for future work toward embodied agents capable of robust causal reasoning rather than correlation driven updates.

This paper investigates the performance of the A* algorithm in the field of automated machine learning using program synthesis. We designed a context-free grammar to create machine learning pipelines and came up with a cost function for A*. Two different experiments were done, the first one to tune the parameters of our algorithm and the second one to compare the efficiency of A* with other search algorithms. The results indicate that for the selected datasets, A* did not have better performance, but rather had similar results with the other search algorithms. Nevertheless, more research in this field is needed to find concrete results.