Haunted House: A text-based game for comparing the flexibility of mental models in humans and LLMs
This work highlights a critical limitation in current LLMs for tasks requiring active model-based reasoning, providing a benchmark for future AI development.
The study introduced a text-based game to compare model-based reasoning in humans and LLMs, finding that humans achieved a 31.6% success rate, significantly outperforming seven state-of-the-art LLMs, with only one successful attempt by Claude 3 Opus out of 140 attempts.
This study introduces "Haunted House" a novel text-based game designed to compare the performance of humans and large language models (LLMs) in model-based reasoning. Players must escape from a house containing nine rooms in a 3x3 grid layout while avoiding the ghost. They are guided by verbal clues that they get each time they move. In Study 1, the results from 98 human participants revealed a success rate of 31.6%, significantly outperforming seven state-of-the-art LLMs tested. Out of 140 attempts across seven LLMs, only one attempt resulted in a pass by Claude 3 Opus. Preliminary results suggested that GPT o3-mini-high performance might be higher, but not at the human level. Further analysis of 29 human participants' moves in Study 2 indicated that LLMs frequently struggled with random and illogical moves, while humans exhibited such errors less frequently. Our findings suggest that current LLMs encounter difficulties in tasks that demand active model-based reasoning, offering inspiration for future benchmarks.