RealFeedApp LLM pricing agents sustain supracompetitive regimes in self-play
The study demonstrates that language model agents can sustain supracompetitive price regimes in repeated Bertrand competition, unlike standard online learners. Their performance advantage against full-information learners vanishes when agents post prices based on summarized market histories.
- Field
- how neural networks work
- What they did
- Researchers demonstrated that language model (LLM) agents in competitive pricing can achieve prices above marginal cost, unlike standard learning algorithms.
- Why it matters
- This is important for understanding how neural networks specifically affect market prices, as their ability to adapt faster allows them to maintain higher prices under certain conditions.
#llm#pricing#bertrand competition#algorithmic game theory#agent learning
