Computer Science & Game Theory

Jiacheng Wang, Jinbin Huang

This provides a foundational theoretical framework explaining reward hacking across AI alignment methods, with implications for all ML/AI systems.

Michał P. Karpowicz

This provides a foundational limit for all of ML/AI, showing that hallucination control is fundamentally impossible, requiring trade-off optimization rather than elimination.

Siqi Zhu, David Zhang, Pedro Cisneros-Velarde et al.

This addresses the issue of misaligned LLM behavior for users in practical applications, offering a novel approach to enhance cooperative outcomes.

Suyash Mishra

This addresses the problem of efficient vulnerability discovery in operating systems for security researchers, presenting a novel integration of game theory and multi-agent orchestration.

Luise Ge, Daniel Halpern, Evi Micha et al. · harvard

This work addresses AI alignment for researchers and practitioners by providing a new axiomatic framework to improve reward learning, though it is incremental as it builds on social choice theory.

Ioannis Anagnostides, Gabriele Farina, Tuomas Sandholm et al.

It solves a foundational optimization problem with broad implications in game theory and nonsmooth optimization, representing a major theoretical advance.

Antoine Scheid, Daniil Tiapkin, Etienne Boursier et al.

This addresses incentive design in applications like healthcare or taxation, extending bandit problems to include learning aspects often overlooked in mechanism design.

Kai Cui, Gökçe Dayanıklı, Mathieu Laurière et al.

This work addresses the problem of modeling multi-player games with major players for researchers in game theory and AI, representing a novel extension rather than an incremental improvement.

Mohammad Ghazi Vakili, Christoph Gorgulla, AkshatKumar Nigam et al.

This work addresses the problem of accelerating drug discovery for cancer patients by demonstrating a novel quantum-assisted approach that yields viable therapeutic candidates.

Lee Cohen, Jack Hsieh, Connie Hong et al. · stanford

This addresses fairness and accuracy issues in hiring for employers and candidates, offering a novel solution to mitigate disparities from differential access to AI tools.

Daniel Ngo, Keegan Harris, Anish Agarwal et al.

This addresses a fundamental issue in causal inference for researchers and practitioners in economics and policy analysis, offering a novel solution to a previously overlooked assumption.

Davide Legacci, Panayotis Mertikopoulos, Christos H. Papadimitriou et al.

This provides foundational insights for multi-agent learning in non-cooperative settings, extending beyond prior work on 2-player zero-sum games.

Raman Ebrahimi, Kristen Vaccaro, Parinaz Naghizadeh

This addresses the challenge of designing AI systems that account for human cognitive biases in strategic settings, representing a novel extension rather than an incremental improvement.

Hanrui Zhang, Yu Cheng, Vincent Conitzer

This work addresses equilibrium computation for game theory and AI applications, offering the first polynomial-time SEFCE algorithm for a general class of stochastic games and the first EFCE algorithm achieving three key desiderata simultaneously, representing a significant advance over prior methods.

Rachitesh Kumar, Jon Schneider, Balasubramanian Sivan

This addresses the need for robust bidding algorithms in online advertising markets, offering the first simultaneous optimal regret and strategic guarantees, though it builds on existing work in game theory and machine learning.

Caner Gocmen, Thodoris Lykouris, Deeksha Sinha et al.

This addresses scheduling inefficiencies in content moderation for social media platforms, offering a novel solution to a domain-specific bottleneck.

Rohan Ghuge, Sahil Singla, Yifan Wang

This work addresses a central problem in computer science, operations research, and economics by providing a more sample-efficient and robust solution, resolving an open question from prior research.

Larkin Liu, Kashif Rasul, Yutong Chao et al.

This work addresses the problem of multi-agent learning in Stackelberg games, which is significant for applications in domains such as cybersecurity and economic supply chain optimization.

Jason Milionis, Jens Ernstberger, Joseph Bonneau et al.

This addresses verification challenges in decentralized networks like DePIN, where untrusted parties provide physical services, offering a theoretical foundation for incentive-compatible mechanisms.

Ermis Soumalias, Yanchen Jiang, Kehang Zhu et al.

This work addresses the problem of efficient preference elicitation for combinatorial assignment, which is significant for applications where human preferences need to be accurately captured, such as course allocation.