Searcharxiv⌕ Search

arXiv subjects

Moshe Tennenholtz

Publications and source records attributed to Moshe Tennenholtz.

At least 55 records · Page 3Linked to original sources
arXiv2023

Optimal Mechanism Design for Agents with DSL Strategies: The Case of Sybil Attacks in Combinatorial Auctions

In robust decision making under uncertainty, a natural choice is to go with safety (aka security) level strategies. However, in many important cases, most notably auctions, there is a large multitude of safety level strategies, thus making the choice unclear. We consider two refined notions: (i) a term we call DSL (distinguishable safety level), and is based on the notion of ``discrimin'', which uses a pairwise comparison of actions while removing trivial equivalencies. This captures the fact that when comparing two actions an agent should not care about payoffs in situations where they lead to identical payoffs. (ii) The well-known Leximin notion from social choice theory, which we apply for robust decision-making. In particular, the leximin is always DSL but not vice-versa. We study the relations of these notions to other robust notions, and illustrate the results of their use in auctions and other settings. Economic design aims to maximize social welfare when facing self-motivated participants. In online environments, such as the Web, participants' incentives take a novel form originating from the lack of clear agent identity -- the ability to create Sybil attacks, i.e., the ability of each participant to act using multiple identities. It is well-known that Sybil attacks are a major obstacle for welfare-maximization. Our main result proves that when DSL attackers face uncertainty over the auction's bids, the celebrated VCG mechanism is welfare-maximizing even under Sybil attacks. Altogether, our work shows a successful fundamental synergy between robustness under uncertainty, economic design, and agents' strategic manipulations in online multi-agent systems.

econ.TH↗
arXiv2023

Selling Data to a Competitor (Extended Abstract)

We study the costs and benefits of selling data to a competitor. Although selling all consumers' data may decrease total firm profits, there exist other selling mechanisms -- in which only some consumers' data is sold -- that render both firms better off. We identify the profit-maximizing mechanism, and show that the benefit to firms comes at a cost to consumers. We then construct Pareto-improving mechanisms, in which each consumers' welfare, as well as both firms' profits, increase. Finally, we show that consumer opt-in can serve as an instrument to induce firms to choose a Pareto-improving mechanism over a profit-maximizing one.

cs.GT↗
arXiv2023

Resilient Information Aggregation

In an information aggregation game, a set of senders interact with a receiver through a mediator. Each sender observes the state of the world and communicates a message to the mediator, who recommends an action to the receiver based on the messages received. The payoff of the senders and of the receiver depend on both the state of the world and the action selected by the receiver. This setting extends the celebrated cheap talk model in two aspects: there are many senders (as opposed to just one) and there is a mediator. From a practical perspective, this setting captures platforms in which strategic experts advice is aggregated in service of action recommendations to the user. We aim at finding an optimal mediator/platform that maximizes the users' welfare given highly resilient incentive compatibility requirements on the equilibrium selected: we want the platform to be incentive compatible for the receiver/user when selecting the recommended action, and we want it to be resilient against group deviations by the senders/experts. We provide highly positive answers to this challenge, manifested through efficient algorithms.

econ.TH↗
arXiv2023

Selling Data to a Competitor

We study the costs and benefits of selling data to a competitor. Although selling all consumers' data may decrease total firm profits, there exist other selling mechanisms -- in which only some consumers' data is sold -- that render both firms better off. We identify the profit-maximizing mechanism, and show that the benefit to firms comes at a cost to consumers. We then construct Pareto-improving mechanisms, in which each consumers' welfare, as well as both firms' profits, increase. Finally, we show that consumer opt-in can serve as an instrument to induce firms to choose a Pareto-improving mechanism over a profit-maximizing one.

cs.GT↗
arXiv2022

Mediated Cheap Talk Design (with proofs)

We study an information design problem with two informed senders and a receiver in which, in contrast to traditional Bayesian persuasion settings, senders do not have commitment power. In our setting, a trusted mediator/platform gathers data from the senders and recommends the receiver which action to play. We characterize the set of implementable action distributions that can be obtained in equilibrium, and provide an $O(n \log n)$ algorithm (where $n$ is the number of states) that computes the optimal equilibrium for the senders. Additionally, we show that the optimal equilibrium for the receiver can be obtained by a simple revelation mechanism.

cs.GT↗
arXiv2022

Data Curation from Privacy-Aware Agents

A data curator would like to collect data from privacy-aware agents. The collected data will be used for the benefit of all agents. Can the curator incentivize the agents to share their data truthfully? Can he guarantee that truthful sharing will be the unique equilibrium? Can he provide some stability guarantees on such equilibrium? We study necessary and sufficient conditions for these questions to be answered positively and complement these results with corresponding data collection protocols for the curator. Our results account for a broad interpretation of the notion of privacy awareness.

cs.GT↗
arXiv2022

Pareto-Improving Data-Sharing

We study the effects of data sharing between firms on prices, profits, and consumer welfare. Although indiscriminate sharing of consumer data decreases firm profits due to the subsequent increase in competition, selective sharing can be beneficial. We show that there are data-sharing mechanisms that are strictly Pareto-improving, simultaneously increasing firm profits and consumer welfare. Within the class of Pareto-improving mechanisms, we identify one that maximizes firm profits and one that maximizes consumer welfare.

cs.GT↗
arXiv2022

Predicting Decisions in Language Based Persuasion Games

Sender-receiver interactions, and specifically persuasion games, are widely researched in economic modeling and artificial intelligence. However, in the classic persuasion games setting, the messages sent from the expert to the decision-maker (DM) are abstract or well-structured signals rather than natural language messages. This paper addresses the use of natural language in persuasion games. For this purpose, we conduct an online repeated interaction experiment. At each trial of the interaction, an informed expert aims to sell an uninformed decision-maker a vacation in a hotel, by sending her a review that describes the hotel. While the expert is exposed to several scored reviews, the decision-maker observes only the single review sent by the expert, and her payoff in case she chooses to take the hotel is a random draw from the review score distribution available to the expert only. We also compare the behavioral patterns in this experiment to the equivalent patterns in similar experiments where the communication is based on the numerical values of the reviews rather than the reviews' text, and observe substantial differences which can be explained through an equilibrium analysis of the game. We consider a number of modeling approaches for our verbal communication setup, differing from each other in the model type (deep neural network vs. linear classifier), the type of features used by the model (textual, behavioral or both) and the source of the textual features (DNN-based vs. hand-crafted). Our results demonstrate that given a prefix of the interaction sequence, our models can predict the future decisions of the decision-maker, particularly when a sequential modeling approach and hand-crafted textual features are applied. Further analysis of the hand-crafted textual features allows us to make initial observations about the aspects of text that drive decision making in our setup

cs.AI↗
arXiv2022

Budget-Constrained Reinforcement of Ranked Objects

Commercial entries, such as hotels, are ranked according to score by a search engine or recommendation system, and the score of each can be improved upon by making a targeted investment, e.g., advertising. We study the problem of how a principal, who owns or supports a set of entries, can optimally allocate a budget to maximize their ranking. Representing the set of ranked scores as a probability distribution over scores, we treat this question as a game between distributions. We show that, in the general case, the best ranking is achieved by equalizing the scores of several disjoint score ranges. We show that there is a unique optimal reinforcement strategy, and provide an efficient algorithm implementing it.

cs.GT↗
arXiv2022

Long-term Data Sharing under Exclusivity Attacks

The quality of learning generally improves with the scale and diversity of data. Companies and institutions can therefore benefit from building models over shared data. Many cloud and blockchain platforms, as well as government initiatives, are interested in providing this type of service. These cooperative efforts face a challenge, which we call ``exclusivity attacks''. A firm can share distorted data, so that it learns the best model fit, but is also able to mislead others. We study protocols for long-term interactions and their vulnerability to these attacks, in particular for regression and clustering tasks. We conclude that the choice of protocol, as well as the number of Sybil identities an attacker may control, is material to vulnerability.

cs.CR↗
arXiv2021

Designing an Automatic Agent for Repeated Language based Persuasion Games

Persuasion games are fundamental in economics and AI research and serve as the basis for important applications. However, work on this setup assumes communication with stylized messages that do not consist of rich human language. In this paper we consider a repeated sender (expert) -- receiver (decision maker) game, where the sender is fully informed about the state of the world and aims to persuade the receiver to accept a deal by sending one of several possible natural language reviews. We design an automatic expert that plays this repeated game, aiming to achieve the maximal payoff. Our expert is implemented within the Monte Carlo Tree Search (MCTS) algorithm, with deep learning models that exploit behavioral and linguistic signals in order to predict the next action of the decision maker, and the future payoff of the expert given the state of the game and a candidate review. We demonstrate the superiority of our expert over strong baselines, its adaptability to different decision makers, and that its selected reviews are nicely adapted to the proposed deal.

cs.CL↗
arXiv2021

Coopetition Against an Amazon

This paper studies cooperative data-sharing between competitors vying to predict a consumer's tastes. We design optimal data-sharing schemes both for when they compete only with each other, and for when they additionally compete with an Amazon -- a company with more, better data. We show that simple schemes -- threshold rules that probabilistically induce either full data-sharing between competitors, or the full transfer of data from one competitor to another -- are either optimal or approximately optimal, depending on properties of the information structure. We also provide conditions under which firms share more data when they face stronger outside competition, and describe situations in which this conclusion is reversed.

cs.GT↗
arXiv2021

Driving the Herd: Search Engines as Content Influencers

In competitive search settings such as the Web, many documents' authors (publishers) opt to have their documents highly ranked for some queries. To this end, they modify the documents - specifically, their content - in response to induced rankings. Thus, the search engine affects the content in the corpus via its ranking decisions. We present a first study of the ability of search engines to drive pre-defined, targeted, content effects in the corpus using simple techniques. The first is based on the herding phenomenon - a celebrated result from the economics literature - and the second is based on biasing the relevance ranking function. The types of content effects we study are either topical or touch on specific document properties - length and inclusion of query terms. Analysis of ranking competitions we organized between incentivized publishers shows that the types of content effects we target can indeed be attained by applying our suggested techniques. These findings have important implications with regard to the role of search engines in shaping the corpus.

cs.IR↗
arXiv2021

Worst-case Bounds on Power vs. Proportion in Weighted Voting Games with Application to False-name Manipulation

Weighted voting games apply to a wide variety of multi-agent settings. They enable the formalization of power indices which quantify the coalitional power of players. We take a novel approach to the study of the power of big vs.~small players in these games. We model small (big) players as having single (multiple) votes. The aggregate relative power of big players is measured w.r.t.~their votes proportion. For this ratio, we show small constant worst-case bounds for the Shapley-Shubik and the Deegan-Packel indices. In sharp contrast, this ratio is unbounded for the Banzhaf index. As an application, we define a false-name strategic normal form game where each big player may split its votes between false identities, and study its various properties. Together, our results provide foundations for the implications of players' size, modeled as their ability to split, on their relative power.

cs.GT↗
arXiv2021

VCG Under False-name Attacks: a Bayesian Analysis

VCG is a classical combinatorial auction that maximizes social welfare. However, while the standard single-item Vickrey auction is false-name-proof, a major failure of multi-item VCG is its vulnerability to false-name attacks. This occurs already in the natural bare minimum model in which there are two identical items and bidders are single-minded. Previous solutions to this challenge focused on developing alternative mechanisms that compromise social welfare. We re-visit the VCG auction vulnerability and consider the bidder behavior in Bayesian settings. In service of that we introduce a novel notion, termed the granularity threshold, that characterizes VCG Bayesian resilience to false-name attacks as a function of the bidder type distribution. Using this notion we show a large class of cases in which VCG indeed obtains Bayesian resilience for the two-item single-minded setting.

cs.GT↗
arXiv2021

Protecting the Protected Group: Circumventing Harmful Fairness

Machine Learning (ML) algorithms shape our lives. Banks use them to determine if we are good borrowers; IT companies delegate them recruitment decisions; police apply ML for crime-prediction, and judges base their verdicts on ML. However, real-world examples show that such automated decisions tend to discriminate against protected groups. This potential discrimination generated a huge hype both in media and in the research community. Quite a few formal notions of fairness were proposed, which take a form of constraints a "fair" algorithm must satisfy. We focus on scenarios where fairness is imposed on a self-interested party (e.g., a bank that maximizes its revenue). We find that the disadvantaged protected group can be worse off after imposing a fairness constraint. We introduce a family of \textit{Welfare-Equalizing} fairness constraints that equalize per-capita welfare of protected groups, and include \textit{Demographic Parity} and \textit{Equal Opportunity} as particular cases. In this family, we characterize conditions under which the fairness constraint helps the disadvantaged group. We also characterize the structure of the optimal \textit{Welfare-Equalizing} classifier for the self-interested party, and provide an algorithm to compute it. Overall, our \textit{Welfare-Equalizing} fairness approach provides a unified framework for discussing fairness in classification in the presence of a self-interested party.

cs.GT↗
arXiv2020

PMI-Masking: Principled masking of correlated spans

Masking tokens uniformly at random constitutes a common flaw in the pretraining of Masked Language Models (MLMs) such as BERT. We show that such uniform masking allows an MLM to minimize its training objective by latching onto shallow local signals, leading to pretraining inefficiency and suboptimal downstream performance. To address this flaw, we propose PMI-Masking, a principled masking strategy based on the concept of Pointwise Mutual Information (PMI), which jointly masks a token n-gram if it exhibits high collocation over the corpus. PMI-Masking motivates, unifies, and improves upon prior more heuristic approaches that attempt to address the drawback of random uniform token masking, such as whole-word masking, entity/phrase masking, and random-span masking. Specifically, we show experimentally that PMI-Masking reaches the performance of prior masking approaches in half the training time, and consistently improves performance at the end of training.

cs.LG↗
arXiv2020

Fiduciary Bandits

Recommendation systems often face exploration-exploitation tradeoffs: the system can only learn about the desirability of new options by recommending them to some user. Such systems can thus be modeled as multi-armed bandit settings; however, users are self-interested and cannot be made to follow recommendations. We ask whether exploration can nevertheless be performed in a way that scrupulously respects agents' interests---i.e., by a system that acts as a fiduciary. More formally, we introduce a model in which a recommendation system faces an exploration-exploitation tradeoff under the constraint that it can never recommend any action that it knows yields lower reward in expectation than an agent would achieve if it acted alone. Our main contribution is a positive result: an asymptotically optimal, incentive compatible, and ex-ante individually rational recommendation algorithm.

cs.GT↗