arXiv q-fin· Martin Bichler, Philip Hierhager, Kassian K\"ock, Fabian Pieroth·· 2 days agoAI Score49
How Language Model Agents Are Reshaping Price Competition
How language model agents reshape price competition
AI summary
Research shows that large language model pricing agents can deviate from the repeated Bertrand competition benchmark where prices converge to marginal costs in homogeneous goods markets, with their performance depending on deployment decisions. Whether LLM pricing agents can maintain prices above marginal costs in self-play is directly related to deployment choices: whether they set prices period-by-period based directly on raw market history including competitor prices, or periodically output probability distributions based on summaries of their own profit history.
Source:arXiv q-fin · arxiv.org
Machine LearningLarge Language Model ApplicationsPaper Research#Papers / Research#Large Language Model Applications#Machine Learning#arXiv