This study investigated how the expectations of others (i.e., top-down processes) and actual perceived behavior (i.e., bottom-up processes) influence negotiations during human-agent interactions. Participants took part in several sessions of the ultimatum game; we investigated the bargaining strategies directed toward the computer agent. To investigate the influence of top-down and bottom-up processes on performance, we designed an experiment wherein (1) participants expected their partners were humans or agents, and (2) agents used different types of algorithmic behavior. Results revealed that irrational decisions, which are characteristic of human-human interactions, emerged when participants believed their opponents were human and when opponent behaviors were ambiguous. Further, we found participants adopted different bargaining strategies according to their expectations and the agent's specific algorithmic behavior. We discuss interplay of the two types of cognitive processing in human-agent interaction.