More Works from Our Lab

DiBS-MTL: Transformation-Invariant Multitask Learning with Direction Oracles

The transformation invariance property of DiBS leads to state-of-the-art multitask learning performance.

Under Review

Cooperative Bargaining Games Without Utilities: Mediated Solutions from Direction Oracles

Kushagra Gupta*, Surya Murthy*, Mustafa O. Karabag, Ufuk Topcu, David Fridovich-Keil
Control and Learning for Autonomous Robotics (CLeAR) Lab & Autonomous Systems Group (ASG)
The University of Texas at Austin
NeurIPS 2025
*Indicates Equal Contribution
Code Paper on arXiv Copy BibTEX Citation

A formation control game demonstrating the invariances of DiBS (our core contribution) to monotonic non-affine transformations of agent costs. Such transformations arise naturally in learning applications—especially in reinforcement learning and learning from human preferences.

Abstract

Cooperative bargaining games are widely used to model resource allocation and conflict resolution. Traditional solutions assume the mediator can access agents utility function values and gradients. However, there is an increasing number of settings, such as human AI interactions, where utility values may be inaccessible or incomparable due to unknown, nonaffine transformations. To model such settings, we consider that the mediator has access only to agents most preferred directions, i.e., normalized utility gradients in the decision space. To this end, we propose a cooperative bargaining algorithm where a mediator has access to only the direction oracle of each agent. We prove that unlike popular approaches such as the Nash and Kalai Smorodinsky bargaining solutions, our approach is invariant to monotonic nonaffine transformations, and that under strong convexity and smoothness assumptions, this approach enjoys global asymptotic convergence to Pareto stationary solutions. Moreover, we show that the bargaining solutions found by our algorithm also satisfy the axioms of symmetry and (under slightly stronger conditions) independence of irrelevant alternatives, which are popular in the literature. Finally, we conduct experiments in two domains, multi agent formation assignment and mediated stock portfolio allocation, which validate these theoretic results. All code for our experiments can be found here.

BibTeX

@article{gupta2025cooperative,
  title={Cooperative Bargaining Games Without Utilities: Mediated Solutions from Direction Oracles},
  author={Gupta, Kushagra and Murthy, Surya and Karabag, Mustafa O. and Topcu, Ufuk and Fridovich-Keil, David},
  journal={arXiv preprint arXiv:2505.14817},
  year={2025}
}