In repeated Sender-Receiver interactions with misaligned incentives, agents learn communication strategies and linear contracts, enabling the Sender to improve its rewards while reducing fairness by extracting much of the Receiver's surplus.
This is learned through contextual ban- dits where the context is the contract, and the two arms correspond to accept and reject
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.MA 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Information and Contract Design for Repeated Interactions between Agents with Misaligned Incentives
In repeated Sender-Receiver interactions with misaligned incentives, agents learn communication strategies and linear contracts, enabling the Sender to improve its rewards while reducing fairness by extracting much of the Receiver's surplus.