Shai Vardi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Shai Vardi has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Adam Wierman
California Institute of Technology, Pasadena, California, United States
3 shared publications Adam Wierman
California Institute of Technology, Pasadena, California 91125
2 shared publications- Junjie Qin
Purdue University, West Lafayette, IN, USA
2 shared publications - Vidyanand Choudhary
Paul Merage School of Business, University of California, Irvine, California 92697
2 shared publications - Alexander Teytelboym
Department of Economics, University of Oxford, Oxford OX1 3UQ, United Kingdom
1 shared publication Alexandros Psomas
Department of Computer Science, Purdue University, West Lafayette, Indiana 47907
1 shared publication
Latest Publications
- 11 August 2026
- Published by Elsevier in European Journal of Operational Research
- Vol. 336 (3) , 1084-1093
Pure Nash equilibrium market recommendations
- 11 August 2026
- Published by Elsevier in European Journal of Operational Research
- Vol. 336 (3) , 1084-1093
Pure Nash equilibrium market recommendations
- Open Access
- 3 August 2026
- Published by Oxford University Press (OUP) in PNAS Nexus
- Vol. 5 (8)
Fragile preferences: A deep dive into order effects in large language models
Large language models (LLMs) are increasingly deployed in decision-support systems for high-stakes domains such as hiring and university admissions, where choices often involve selecting among competing alternatives. While prior work has noted position biases in LLM-driven comparisons, these biases have not been systematically analyzed or linked to underlying preference structures. We present the first comprehensive study of position biases across multiple LLMs and two distinct domains: resume comparisons, representing a realistic high-stakes context, and color selection, which isolates position effects by removing confounding factors. We find strong and consistent order effects, including a quality-dependent shift: when all options are high quality, models favor the first option, but when quality is lower, they favor later options. We also identify a previously undocumented bias: a name bias, where certain names are favored despite controlling for demographic signals. To separate superficial tie-breaking from genuine distortions of judgment, we extend the rational choice framework to classify pairwise preferences as robust, fragile, or indifferent. Using this framework, we show that order effects can lead models to select strictly inferior options. These results indicate that LLMs exhibit distinct failure modes not documented in human decision-making. We also propose targeted mitigation strategies, including a novel use of the temperature parameter, to recover underlying preferences when order effects distort model behavior.
- 3 August 2026
- Published by Oxford University Press (OUP) in PNAS Nexus
- Vol. 5 (8)
Open AccessFragile preferences: A deep dive into order effects in large language models
Large language models (LLMs) are increasingly deployed in decision-support systems for high-stakes domains such as hiring and university admissions, where choices often involve selecting among competing alternatives. While prior work has noted position biases in LLM-driven comparisons, these biases have not been systematically analyzed or linked to underlying preference structures. We present the first comprehensive study of position biases across multiple LLMs and two distinct domains: resume comparisons, representing a realistic high-stakes context, and color selection, which isolates position effects by removing confounding factors. We find strong and consistent order effects, including a quality-dependent shift: when all options are high quality, models favor the first option, but when quality is lower, they favor later options. We also identify a previously undocumented bias: a name bias, where certain names are favored despite controlling for demographic signals. To separate superficial tie-breaking from genuine distortions of judgment, we extend the rational choice framework to classify pairwise preferences as robust, fragile, or indifferent. Using this framework, we show that order effects can lead models to select strictly inferior options. These results indicate that LLMs exhibit distinct failure modes not documented in human decision-making. We also propose targeted mitigation strategies, including a novel use of the temperature parameter, to recover underlying preferences when order effects distort model behavior.
- 22 May 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Information Systems Research
A Random Walk Modeling Framework for Boosting the Creativity of Humans and AI
Innovation depends on generating ideas that are not merely more numerous but meaningfully different. This article offers a practical framework for improving creativity in both people and generative artificial intelligence (AI). It models ideation as movement through a network of concepts, where common associations are easy to reach and more creative ideas lie on less-traveled paths. The framework explains why familiar techniques work and gives managers a systematic way to create new ones: narrow attention through constraints, shift the context, or delay output before selecting an answer. Across experiments with human participants and large language models, simple interventions—such as asking for alliterative responses, introducing an unusual environment, or requesting the fifth idea rather than the first—raised the novelty of responses compared with baseline prompts. For practice, the findings show that creativity can be improved through small changes to prompts, workflows, brainstorming rules, and AI-copilot interfaces. For policy, the study suggests that AI adoption should focus not only on model capability, but also on how human-AI interaction is structured to preserve diversity, reduce fixation, and support innovation across organizations.
- 22 May 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Information Systems Research
A Random Walk Modeling Framework for Boosting the Creativity of Humans and AI
Innovation depends on generating ideas that are not merely more numerous but meaningfully different. This article offers a practical framework for improving creativity in both people and generative artificial intelligence (AI). It models ideation as movement through a network of concepts, where common associations are easy to reach and more creative ideas lie on less-traveled paths. The framework explains why familiar techniques work and gives managers a systematic way to create new ones: narrow attention through constraints, shift the context, or delay output before selecting an answer. Across experiments with human participants and large language models, simple interventions—such as asking for alliterative responses, introducing an unusual environment, or requesting the fifth idea rather than the first—raised the novelty of responses compared with baseline prompts. For practice, the findings show that creativity can be improved through small changes to prompts, workflows, brainstorming rules, and AI-copilot interfaces. For policy, the study suggests that AI adoption should focus not only on model capability, but also on how human-AI interaction is structured to preserve diversity, reduce fixation, and support innovation across organizations.
- 1 November 2025
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 73 (6) , 3104-3117
The Price of Fairness of Scheduling a Scarce Resource
Balancing Fairness and Efficiency: The Price of Fairness in Scheduling Scarce Resources “The Price of Fairness of Scheduling a Scarce Resource” investigates the trade-offs between fairness and utility when scheduling scarce resources among multiple stakeholders. The research focuses on envy-freeness, a widely recognized fairness criterion, and explores how fairness constraints impact overall utility, quantified by the price of fairness—the ratio of attainable utility with and without fairness considerations. By analyzing various factors, including the number of agents, time horizon, discount factors, and switching costs, the authors derive tight bounds on the price of fairness. Interestingly, they discover that, whereas fairness often increases utility loss, certain parameters—such as reducing the time horizon or increasing the number of agents—can counterintuitively reduce this loss. These findings offer strategic insights for decision makers facing fairness constraints in scheduling, helping them navigate complex trade-offs between equitable resource allocation and maximizing efficiency.
- 1 November 2025
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 73 (6) , 3104-3117
The Price of Fairness of Scheduling a Scarce Resource
Balancing Fairness and Efficiency: The Price of Fairness in Scheduling Scarce Resources “The Price of Fairness of Scheduling a Scarce Resource” investigates the trade-offs between fairness and utility when scheduling scarce resources among multiple stakeholders. The research focuses on envy-freeness, a widely recognized fairness criterion, and explores how fairness constraints impact overall utility, quantified by the price of fairness—the ratio of attainable utility with and without fairness considerations. By analyzing various factors, including the number of agents, time horizon, discount factors, and switching costs, the authors derive tight bounds on the price of fairness. Interestingly, they discover that, whereas fairness often increases utility loss, certain parameters—such as reducing the time horizon or increasing the number of agents—can counterintuitively reduce this loss. These findings offer strategic insights for decision makers facing fairness constraints in scheduling, helping them navigate complex trade-offs between equitable resource allocation and maximizing efficiency.
- Open Access
- 2 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 476
Efficiency, Envy and Incentives in Combinatorial Assignment
- 2 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 476
Open AccessEfficiency, Envy and Incentives in Combinatorial Assignment
- 1 September 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 72 (5) , 2076-2081
Technical Note—On Hiring Secretaries with Stochastic Departures
The paper studies generalization of the secretary problem, where decisions do not have to be made immediately upon applicants’ arrivals. After arriving, each applicant stays in the system for some (random) amount of time and then leaves, whereupon the algorithm has to decide irrevocably whether to select this applicant or not. The arrival and waiting times are drawn from known distributions, and the decision maker’s goal is to maximize the probability of selecting the best applicant overall. The paper characterizes the optimal policy for this setting, showing that when deciding whether to select an applicant, it suffices to know only the time and the number of applicants that have arrived so far. Furthermore, the policy is monotone nondecreasing in the number of applicants seen so far, and, under certain natural conditions, monotone nonincreasing in time. Furthermore, when the number of applicants is large, a single threshold policy is almost optimal.
- 1 September 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 72 (5) , 2076-2081
Technical Note—On Hiring Secretaries with Stochastic Departures
The paper studies generalization of the secretary problem, where decisions do not have to be made immediately upon applicants’ arrivals. After arriving, each applicant stays in the system for some (random) amount of time and then leaves, whereupon the algorithm has to decide irrevocably whether to select this applicant or not. The arrival and waiting times are drawn from known distributions, and the decision maker’s goal is to maximize the probability of selecting the best applicant overall. The paper characterizes the optimal policy for this setting, showing that when deciding whether to select an applicant, it suffices to know only the time and the number of applicants that have arrived so far. Furthermore, the policy is monotone nondecreasing in the number of applicants seen so far, and, under certain natural conditions, monotone nonincreasing in time. Furthermore, when the number of applicants is large, a single threshold policy is almost optimal.
- Open Access
- 8 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 242
Ex-Post Equilibrium Market Recommendations
New business models help sellers make better decisions by communicating information about market prices. However, sellers' inability to coordinate greatly reduces the efficacy of this information and can lead to market failures. We study the feasibility and benefits of providing equilibria as recommendations in economies where sellers face price uncertainty, information scarcity, and an inability to coordinate. We describe a general model in which sellers wish to sell some quantity of a good at one of several markets with elastic prices, and show that while an equilibrium may not necessarily exist, an approximate equilibrium always exists under mild assumptions on the market concentration. We describe an algorithm that a market planner can use to recommend approximate ex-post equilibria to sellers, and compare the recommended (approximate) equilibria to selling strategies identified from a survey of onion and potato farmers in India. Using both synthetic data and real data on India's agricultural markets, we show that the recommended strategy outperforms the other strategies in terms of seller welfare, geographic price dispersion, market volume concentration, and several metrics of individual farmer welfare improvements. We also show that the framework is robust to imperfect market penetration, is fair to smaller farmers, and captures nearly all of the welfare from a welfare-maximizing allocation.
- 8 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 242
Open AccessEx-Post Equilibrium Market Recommendations
New business models help sellers make better decisions by communicating information about market prices. However, sellers' inability to coordinate greatly reduces the efficacy of this information and can lead to market failures. We study the feasibility and benefits of providing equilibria as recommendations in economies where sellers face price uncertainty, information scarcity, and an inability to coordinate. We describe a general model in which sellers wish to sell some quantity of a good at one of several markets with elastic prices, and show that while an equilibrium may not necessarily exist, an approximate equilibrium always exists under mild assumptions on the market concentration. We describe an algorithm that a market planner can use to recommend approximate ex-post equilibria to sellers, and compare the recommended (approximate) equilibria to selling strategies identified from a survey of onion and potato farmers in India. Using both synthetic data and real data on India's agricultural markets, we show that the recommended strategy outperforms the other strategies in terms of seller welfare, geographic price dispersion, market volume concentration, and several metrics of individual farmer welfare improvements. We also show that the framework is robust to imperfect market penetration, is fair to smaller farmers, and captures nearly all of the welfare from a welfare-maximizing allocation.
- 1 May 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Mathematics of Operations Research
- Vol. 49 (2) , 928-947
Minimization Fractional Prophet Inequalities for Sequential Procurement
We consider a minimization variant on the classical prophet inequality with monomial cost functions. A firm would like to procure some fixed amount of a divisible commodity from sellers that arrive sequentially. Whenever a seller arrives, the seller’s cost function is revealed, and the firm chooses how much of the commodity to buy. We first show that if one restricts the set of distributions for the coefficients to a family of natural distributions that include, for example, the uniform and truncated normal distributions, then there is a thresholding policy that is asymptotically optimal in the number of sellers. We then compare two scenarios based on whether the firm has in-house production capabilities or not. We precisely compute the optimal algorithm’s competitive ratio when in-house production capabilities exist and for a special case when they do not. We show that the main advantage of the ability to produce the commodity in house is that it shields the firm from price spikes in worst-case scenarios. Funding: This work was supported by NSF Grants [CNS-2146814, CPS-2136197, CNS-2106403, NGSDI-2105648].
- 1 May 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Mathematics of Operations Research
- Vol. 49 (2) , 928-947
Minimization Fractional Prophet Inequalities for Sequential Procurement
We consider a minimization variant on the classical prophet inequality with monomial cost functions. A firm would like to procure some fixed amount of a divisible commodity from sellers that arrive sequentially. Whenever a seller arrives, the seller’s cost function is revealed, and the firm chooses how much of the commodity to buy. We first show that if one restricts the set of distributions for the coefficients to a family of natural distributions that include, for example, the uniform and truncated normal distributions, then there is a thresholding policy that is asymptotically optimal in the number of sellers. We then compare two scenarios based on whether the firm has in-house production capabilities or not. We precisely compute the optimal algorithm’s competitive ratio when in-house production capabilities exist and for a special case when they do not. We show that the main advantage of the ability to produce the commodity in house is that it shields the firm from price spikes in worst-case scenarios. Funding: This work was supported by NSF Grants [CNS-2146814, CPS-2136197, CNS-2106403, NGSDI-2105648].
- 1 March 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 72 (2) , 796-815
Black-Box Acceleration of Monotone Convex Program Solvers
When and where was the study conducted: This work was done in 2018, 2019 and 2020 when Palma London was a PhD student at Caltech and Shai Vardi was a postdoc at Caltech. This work was also done in part while Palma London was visiting Purdue University, and while Reza Eghbali was a postdoctoral fellow the Simons Institute for the Theory of Computing. Adam Wierman is a professor at Caltech. Article Summary and Talking Points: Please describe the primary purpose/findings of your article in 3 sentences or less. This paper presents a framework for accelerating (speeding up) existing convex program solvers. Across engineering disciplines, a fundamental bottleneck is the availability of fast, efficient, accurate solvers. We present an acceleration method that speeds up linear programing solvers such as Gurobi and convex program solvers such as the Splitting Conic Solver by two orders of magnitude. Please include 3-5 short bullet points of “Need to Know” items regarding this research and your findings. - Optimizations problems arise in many engineering and science disciplines, and developing efficient optimization solvers is key to future innovation. - We speed up linear programing solver Gurobi by two orders of magnitude. - This work applies to optimization problems with monotone objective functions and packing constraints, which is a common problem formulation across many disciplines. Please identify 2 pull quotes from your article that best capture the novelty and impact of your research. “We propose a framework for accelerating exact and approximate convex programming solvers for packing linear programming problems and a family of convex programming problems with linear constraints. Analytically, we provide worst-case guarantees on the run time and the quality of the solution produced. Numerically, we demonstrate that our framework speeds up Gurobi and the Splitting Conic Solver by two orders of magnitude, while maintaining a near-optimal solution.” “Our focus in this paper is on a class of packing problems for which data is either very costly or hard to obtain. In these situations, the number of data points available is much smaller than the number of variables. In a machine-learning setting, this regime is increasingly prevalent because it is often advantageous to consider larger and larger feature spaces, while not necessarily obtaining proportionally more data.” Article Implications - Please describe in 5 sentences or less the innovative takeaway(s) of your research. This framework applies to optimization problems with monotone objective functions and packing constraints, which is a common problem formulation across many disciplines, including machine learning, inference, and resource allocation. Providing fast solvers for these problems is crucial. We exploit characteristics of the problem structure and leverage statistical properties of the problem constraints to allow us to speed up optimization solvers. We present worst-case guarantees on run-time, and empirically demonstrate speedups of two orders of magnitude. - Please describe in 5 sentences or less why your findings would be of interest to the general public. Many problems in engineering, science, math, and machine learning involve solving an optimization problem. Fast, efficient optimization solvers are key to future innovation in science and engineering. This work presents a tool to accelerate existing convex solvers, and thus can also be applied to future solvers. As the size of datasets grow it is even more crucial to have fast solvers. - Who would be the most impacted by your research (i.e. by industry, job title, consumer category). Our work impacts machine-learning researchers and optimization researchers, in industry or academia.
- 1 March 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 72 (2) , 796-815
Black-Box Acceleration of Monotone Convex Program Solvers
When and where was the study conducted: This work was done in 2018, 2019 and 2020 when Palma London was a PhD student at Caltech and Shai Vardi was a postdoc at Caltech. This work was also done in part while Palma London was visiting Purdue University, and while Reza Eghbali was a postdoctoral fellow the Simons Institute for the Theory of Computing. Adam Wierman is a professor at Caltech. Article Summary and Talking Points: Please describe the primary purpose/findings of your article in 3 sentences or less. This paper presents a framework for accelerating (speeding up) existing convex program solvers. Across engineering disciplines, a fundamental bottleneck is the availability of fast, efficient, accurate solvers. We present an acceleration method that speeds up linear programing solvers such as Gurobi and convex program solvers such as the Splitting Conic Solver by two orders of magnitude. Please include 3-5 short bullet points of “Need to Know” items regarding this research and your findings. - Optimizations problems arise in many engineering and science disciplines, and developing efficient optimization solvers is key to future innovation. - We speed up linear programing solver Gurobi by two orders of magnitude. - This work applies to optimization problems with monotone objective functions and packing constraints, which is a common problem formulation across many disciplines. Please identify 2 pull quotes from your article that best capture the novelty and impact of your research. “We propose a framework for accelerating exact and approximate convex programming solvers for packing linear programming problems and a family of convex programming problems with linear constraints. Analytically, we provide worst-case guarantees on the run time and the quality of the solution produced. Numerically, we demonstrate that our framework speeds up Gurobi and the Splitting Conic Solver by two orders of magnitude, while maintaining a near-optimal solution.” “Our focus in this paper is on a class of packing problems for which data is either very costly or hard to obtain. In these situations, the number of data points available is much smaller than the number of variables. In a machine-learning setting, this regime is increasingly prevalent because it is often advantageous to consider larger and larger feature spaces, while not necessarily obtaining proportionally more data.” Article Implications - Please describe in 5 sentences or less the innovative takeaway(s) of your research. This framework applies to optimization problems with monotone objective functions and packing constraints, which is a common problem formulation across many disciplines, including machine learning, inference, and resource allocation. Providing fast solvers for these problems is crucial. We exploit characteristics of the problem structure and leverage statistical properties of the problem constraints to allow us to speed up optimization solvers. We present worst-case guarantees on run-time, and empirically demonstrate speedups of two orders of magnitude. - Please describe in 5 sentences or less why your findings would be of interest to the general public. Many problems in engineering, science, math, and machine learning involve solving an optimization problem. Fast, efficient optimization solvers are key to future innovation in science and engineering. This work presents a tool to accelerate existing convex solvers, and thus can also be applied to future solvers. As the size of datasets grow it is even more crucial to have fast solvers. - Who would be the most impacted by your research (i.e. by industry, job title, consumer category). Our work impacts machine-learning researchers and optimization researchers, in industry or academia.
- 13 November 2023
- Published by Elsevier in Operations Research Letters
- Vol. 51 (6) , 709-716
Communication-aware scheduling of precedence-constrained tasks on related machines
- 13 November 2023
- Published by Elsevier in Operations Research Letters
- Vol. 51 (6) , 709-716
Communication-aware scheduling of precedence-constrained tasks on related machines