Science can be viewed as a collective, epistemic endeavor. However, a variety of factors- such as the publish-or-perish culture, institutional incentives, and publishers who favor novel and positive findings- may challenge the ability of science to accurately aggregate information about the world. Evidence of the shortcomings in the current structure of science can be seen in the replication crisis that faces psychology and other disciplines. We analyze scientific publishing through the lens of cultural evolution, framing the scientific process as a multi-generational interplay between scientists and publishers in a multi-armed bandit setting. We examine the dynamics of this model through simulations, exploring the effect that different publication policies have on the accuracy of the published scientific record. Our findings highlight the need for replications and caution against behaviors that prioritize factors uncorrelated with result accuracy.

We are able to learn from others through a combination of trust and vigilance: we trust and believe people who are reliable and have our interests at heart; we ignore those who are incompetent or self-interested. While past work has studied how others' competence influences social learning, relatively little attention has been paid to how others' motivations influence such processes. To address this gap, we develop a Bayesian model of vigilance that considers the speaker's instrumental self-interest, and test predictions of this model through an experiment. In accordance with our model, participants become more vigilant when informants stand to benefit from influencing their actions. When perceived self-interest is maximal, testimony can be discounted wholesale, rendering middle ground increasingly difficult, if not impossible, to find. Our results have implications for research on polarization, misinformation, and disagreement.

Speakers communicate to influence their partner's beliefs and shape their actions. Belief- and action-based objectives have been explored independently in recent computational models, but it has been challenging to explicitly compare or integrate them. Indeed, we find that they are conflated in standard referential communication tasks. To distinguish these accounts, we introduce a new paradigm called signaling bandits, generalizing classic Lewis signaling games to a multi-armed bandit setting where all targets in the context have some relative value. We develop three speaker models: a belief-oriented speaker with a purely informative objective; an action-oriented speaker with an instrumental objective; and a combined speaker which integrates the two by inducing listener beliefs that generally lead to desirable actions. We then present a series of simulations demonstrating that grounding production choices in future listener actions results in relevance effects and flexible uses of nonliteral language. More broadly, our findings suggest that language games based on richer decision problems are a promising avenue for insight into rational communication.

Our aspirations are influenced by the rewards obtained by people around us. How adaptive are these inherited aspirations in stochastic, lottery-like environments? We study the behavior of social Q-learning agents in two multi-armed bandit (MAB) settings: 1) a standard task where one arm gives a higher reward than others and, 2) a lottery task where all arms give a high reward with some small probability. We define aspiration as a function of rewards attained by a previous generation, and happiness as a linear combination of rewards and aspiration. We find that in the standard MAB task, higher aspiration encourages exploration, and agents who learn from the ‘top’ agents accumulate more rewards and happiness. However, in the lottery task, higher aspiration doesn’t improve performance; instead, agents who learn from the ‘top’ agents are more unhappy. Together, this research highlights the context-dependent nature of aspirations and their implications to modern society.

Similarity judgments provide a well-established method for accessing mental representations, with applications in psychology, neuroscience and machine learning. However, collecting similarity judgments can be prohibitively expensive for naturalistic datasets as the number of comparisons grows quadratically in the number of stimuli. One way to tackle this problem is to construct approximation procedures that rely on more accessible proxies for predicting similarity. Here we leverage recent advances in language models and online recruitment, proposing an efficient domain-general procedure for predicting human similarity judgments based on text descriptions. Intuitively, similar stimuli are likely to evoke similar descriptions, allowing us to use description similarity to predict pairwise similarity judgments. Crucially, the number of descriptions required grows only linearly with the number of stimuli, drastically reducing the amount of data required. We test this procedure on six datasets of naturalistic images and show that our models outperform previous approaches based on visual information.