Modeling Bounded Rationality of Agents During Interactions