LLM-driven Imitation of Subrational Behavior : Illusion or Reality?
By Andrea Coletta, Kshama Dwarakanath, Penghang Liu, Svitlana Vyetrenko, Tucker Balch
Rating
1389
Battle Count: 20
Relevance
7/10
The framework could be adapted to model investor behavior in financial markets, potentially improving agent-based models of market dynamics and trading strategies that account for human irrationality.
Implementation Complexity
6/10
While the basic framework is straightforward, careful prompt engineering and integration with LLMs may require significant effort. Adapting to specific trading scenarios could increase complexity.
Reproducibility
4/5
The paper provides detailed experimental settings and prompts used for LLM demonstrations, enhancing reproducibility. However, the exact version of GPT-4 used may affect results.
About this paper
Methodology: LLM-driven Imitation Learning. Problem types: Imitation Learning, Behavioral Modeling, Decision Making.
The interactive Everscope explorer (charts, battles, favorites) loads below.