Max Kleiman-Weiner

Max Kleiman-Weiner

Cognitive Science + AI

Harvard / MIT

I am a Computational Cognitive Scientist affiliated with Harvard SEAS and MIT BCS. My interests span the study of intelligence (natural), intelligence (artificial), intelligence (collective), and intelligence (moral). My goal is to create computational models that explain how the mind works and draw on insights from how people learn and think to build smarter and more human-like artificial intelligence. I focus on social intelligence and have worked on:

  • Social Learning: How do we infer the hidden contents of other people’s minds and learn what they know, want, and think? Or distinguish friend and foe based on just a few social interactions? How do we “download” knowledge from a cumulative culture or “upload” our discoveries through teaching and communication? Can we build machines with “Theory-of-Mind” that learn from people as quickly and intuitively as we do?
  • Cooperation: When should we cooperate with others and when should we compete? How do we coordinate on shared goals to achieve together what none of us could achieve on our own? Can we build machines that collaborate with us as flexibly and fairly as a friend or colleague?
  • Morality: What are moral values, how do we learn them so quickly, and why do we generalize them so widely? Where do our moral values come from and where are they going? How should we build moral machines? Will they amplify our biases or help us overcome them?

Recent breakthroughs in cognitive science, made possible by new tools from machine learning, are allowing us to give formal answers to these questions for the first time. To this end, I build models that integrate the best features of probabilistic causal models, deep and reinforcement learning, and evolutionary game theory. These models give precise accounts of human social cognition and make fine-grained predictions that I test empirically in multi-player behavioral experiments.

I earned my PhD in Computational Cognitive Science from MIT, advised by Josh Tenenbaum, where I received fellowships from the Hertz Foundation and NSF. Previously, I was a Marshall Scholar in Statistics at Oxford advised by Tim Behrens, a Fulbright Fellow in Beijing with Scott Rozelle, and before that was an undergraduate at Stanford where I worked with John Huguenard. Finally, I co-founded two companies: Common Sense Machines and Diffeo. I grew up in Santa Monica and outside of science, I enjoy surfing, skiing, and sushi.


(2023). Learning to Coordinate with Humans using Action Features. International Conference on Machine Learning. PDF Cite

(2023). Virtue Discounting: Observers Infer That Publicly Virtuous Actors Have Less Principled Motivations. Open Mind. Cite

(2022). Assessing and dissociating virtues from the ‘bottom up’: A case study of generosity vs. fairness. The Journal of Positive Psychology. PDF Cite

(2022). Does big data serve policy? Not without context. An experiment with in silico social science. Computational and Mathematical Organization Theory. PDF Cite

(2022). Overloaded communication as paternalistic helping. Proceedings of the Annual Meeting of the Cognitive Science Society. PDF Cite

(2020). Drivers are blamed more than their automated cars when both make mistakes. Nature human behaviour. PDF Cite

(2020). Nonlinearities between inhibition and T-type calcium channel activity bidirectionally regulate thalamic oscillations. Elife. PDF Cite

(2020). The logic of universalization guides moral judgment. Proceedings of the National Academy of Sciences. PDF Cite

(2020). What we owe to family: The impact of special obligations on moral judgment. Psychological Science. PDF Cite

(2018). A Computational Model of Commonsense Moral Decision Making. Proceedings of AAAI/ACM Artificial Intelligence, Ethics and Society (AIES) conference. PDF Cite

(2018). Learning to share and hide intentions using information regularization. Advances in Neural Information Processing Systems. PDF Cite

(2018). Lucky or clever? From expectations to responsibility judgments. Cognition. PDF Cite

(2018). Non-parametric Bayesian Inference of Strategies in Repeated Games. The Econometrics Journal. PDF Cite

(2018). People Make the Same Bayesian Judgment They Criticize in Others. Psychological Science. PDF Cite

(2018). The Cognitive Mechanisms of Contractualist Moral Decision-Making. Proceedings of the Annual Meeting of the Cognitive Science Society. PDF Cite

(2018). Thesis: Computational foundations of human social intelligence. PDF Cite

(2018). Towards Formal Definitions of Blameworthiness, Intention, and Moral Responsibility. Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence (AAAI-18). PDF Cite

(2017). Statistically inaccurate and morally unfair judgements via base rate intrusion. Nature Human Behaviour. PDF Cite

(2016). Feature-based joint planning and norm learning in collaborative games. Proceedings of the 38th Annual Conference of the Cognitive Science Society. PDF Cite

(2015). Go fishing! Responsibility judgments when cooperation breaks down. Proceedings of the 37th Annual Conference of the Cognitive Science Society. PDF Cite

(2014). Wins above replacement. Proceedings of the 36th Annual Conference of the Cognitive Science Society. PDF Cite

(2013). Eggs versus chewable vitamins: Which intervention can increase nutrition and test scores in rural China?. China Economic Review. PDF Cite

(2013). Evaluating Stream Filtering for Entity Profile Updates for TREC 2013 (KBA Track Overview). Proceedings of the Text REtrieval Conference (TREC). PDF Cite

(2013). Multiple micronutrient supplementation reduces anemia and anxiety in rural China’s elementary school children. The Journal of Nutrition. PDF Cite

(2012). Building an entity-centric stream filtering test collection for TREC 2012. Proceedings of the Text REtrieval Conference (TREC). PDF Cite

(2010). Anemia in rural China's elementary schools: Prevalence and correlates in Shaanxi province's poor counties. Ecology of Food and Nutrition. PDF Cite

(2010). Rescuing the corticostriatal synaptic disconnection in the R6/2 mouse model of Huntington's disease: exercise, adenosine receptors and ampakines. PLoS Currents. PDF Cite

(2009). A gain in GABAA receptor synaptic strength in thalamus reduces oscillatory activity and absence seizures. Proceedings of the National Academy of Sciences. PDF Cite

(2009). Synergistic roles of GABAA receptors and SK channels in regulating thalamocortical oscillations. Journal of Neurophysiology. PDF Cite

(2008). Differential electrophysiological properties of dopamine D1 and D2 receptor-containing striatal medium-sized spiny neurons. European Journal of Neuroscience. PDF Cite

(2006). The sound of one arm swinging: a model for multidimensional auditory display of physical motion. Proceedings of the 14th International Conference on Auditory Display. PDF Cite