Abstract: Clinical prediction models often encode group attributes like sex, age, and HIV status for personalization – i.e., to assign more accurate predictions to heterogeneous subpopulations. In this talk, I will describe how such practices inadvertently lead to worsenalization, by assigning unnecessarily inaccurate predictions to minority groups. I will discuss how these effects violate our basic expectations from personalization...
Abstract: What happens when machine learning is outsourced to profit-maximizing agents?
In this work, we propose a theoretical framework for incentive-aware delegation of machine learning tasks. We model delegation as a principal-agent game, in which accurate learning can be incentivized by the principal using performance-based...
Abstract: We tackle the problem of Miners Extracted Value (MEV), where a transaction sequencer extracts profit off users, by frontrunning pending transactions, sandwiching others, or carrying out any other form of ordering manipulations. Most transaction settlement services are designed such that, in each round, a single sequencer is granted full, unrestricted permission to sequence pending transactions. Accordingly, to undermine sequencer monopoly, we devise an auction that forces competition...
Title: Reinforcement Learning Meets Bilevel Optimization: Learning Leader-Follower Games with Sample Efficiency
Abstract: In this talk, I will introduce methods that modify the optimism principle for reinforcement learning in leader-follower games, especially when the follower's reward function is unknown. Such problems generally face statistical challenges due to the ill-posed nature of the best response function. I will discuss two cases that overcome these challenges. The first involves a fully rational follower with a...
Title: Reveal Planning Capability of Autoregressive Learning in Large Language Models
Abstract: Search and planning are fundamental constructs of human intelligence, involved in almost every aspect of our daily lives, from completing tasks at work to organizing trips, to seeking mathematical proofs of theorems, and more. Studying the planning capabilities of large language models (LLMs) can help us understand the differences in the decision-making processes between humans and artificial...
Abstract: We study two combinatorial settings of the contract design problem, in which a principal wants to delegate the execution of a costly task. In the first setting, the principal delegates the task to an agent that can take any subset of a given set of unobservable actions, each of which has an associated cost. The principal receives a reward which is a combinatorial function of the actions taken by the agent. In the second setting, we study the single-principal multi-...
Abstract: Equilibrium problems in Bayesian auction games can be described as systems of differential equations. Depending on the model assumptions, these equations might be such that we do not have an exact mathematical solution theory. The lack of analytical or numerical techniques with guaranteed convergence for the equilibrium problem has plagued the field and limited equilibrium analysis to rather simple auction models such as single-object auctions. Recent progress in equilibrium learning led to algorithms that find approximate...
Title: Computing Optimal Equilibria and Mechanisms via Learning in Zero-Sum Extensive-Form Games
Abstract: We introduce a unified framework for understanding optimal equilibria in general games by explicitly augmenting the game with a mediator who can communicate with players. By varying restrictions governing the communication, the framework covers a wide family of solution concepts and problems including---but not limited to---communication...
Title: Strategic Classification: Learning With Data That 'Behaves'
Abstract: The growing success of machine learning across a wide range of domains and applications has made it appealing to be used also as a tool for informing decisions about humans. But humans are not your conventional input: they have goals, beliefs, and aspirations, and take action to promote their own self-interests. Given that standard learning methods are not designed to handle inputs that "behave", it is natural to ask: how should we...