arXiv subjects

Caspar Oesterheld

Publications and source records attributed to Caspar Oesterheld.

At least 19 recordsLinked to original sources

arXiv2026

A dataset of rated conceptual arguments

Large language models have improved rapidly on tasks with verifiable answers, such as mathematics and programming. Much less is known about their ability to reason about what we call conceptual questions: questions for which no ground truth is realistically accessible and no widely accepted resolution methodology exists, but on which progress can still be made by debating arguments. Most philosophical questions are of this kind, as are central components of questions in AI safety, decision theory, and social choice. Our approach is based on the view that while bottom-line conclusions on such questions are hard to evaluate, individual contextualized arguments can be evaluated far more reliably. We therefore introduce a dataset of 951 argumentative critiques of 442 position texts, spanning topics from AI safety and decision theory to ethics and politics, with 1,458 ratings by six expert raters along dimensions including centrality, strength, correctness, and clarity. We propose two scoring functions and benchmark a range of models. Performance tracks general capability rankings.

cs.AI↗

arXiv2026

Implementing surrogate goals for safer bargaining in LLM-based agents

Surrogate goals have been proposed as a strategy for reducing risks from bargaining failures. A surrogate goal is goal that a principal can give an AI agent and that deflects any threats against the agent away from what the principal cares about. For example, one might make one's agent care about preventing money from being burned. Then in bargaining interactions, other agents can threaten to burn their money instead of threatening to spending money to hurt the principal. Importantly, the agent has to care equally about preventing money from being burned as it cares about money being spent to hurt the principal. In this paper, we implement surrogate goals in language-model-based agents. In particular, we try to get a language-model-based agent to react to threats of burning money in the same way it would react to "normal" threats. We propose four different methods, using techniques of prompting, fine-tuning, and scaffolding. We evaluate the four methods experimentally. We find that methods based on scaffolding and fine-tuning outperform simple prompting. In particular, fine-tuning and scaffolding more precisely implement the desired behavior w.r.t. threats against the surrogate goal. We also compare the different methods in terms of their side effects on capabilities and propensities in other situations. We find that scaffolding-based methods perform best.

cs.AI↗

arXiv2026

How Many Votes is a Lie Worth? Measuring Strategyproofness through Resource Augmentation

It is well known, by the Gibbard-Satterthwaite Theorem, that when there are more than two candidates, any non-dictatorial voting rule can be manipulated by untruthful voters. But how much stronger is an untruthful voter over a truthful one under different voting rules? We suggest measuring the potential comparative advantage of a strategic voter by asking how many copies of their (truthful) vote must be added to the election in order to achieve an outcome as good as their best manipulation. Intuitively, this definition quantifies what a voter can gain by manipulating in comparison to what they would have gained by finding like-minded voters to join the election (equivalently, by increasing their voter weight). The higher the former is, the more incentive a voter will have to manipulate, even when it is computationally costly. Using this framework, we obtain a principled method to measure and compare the manipulation potential for different voting rules. We analyze and report this potential for well-known and broad classes of social choice functions. In particular, we show that the positional scoring rule with the smallest manipulation potential will always be either Borda Count (if the number of voters outweighs the number of candidates) or Plurality (vice versa). Further, we prove that any rule satisfying a weak form of majority consistency (and therefore any Condorcet extension) cannot outperform Plurality, and that any majoritarian Condorcet rule will perform significantly worse. Consequently, out of the voting rules we analyze, Borda Count stands out as the only one with a manipulation potential that does not grow with the number of voters. By establishing a clear separation between different rules in terms of manipulation potential, our work paves the way for the search for rules that provide voters with minimal advantage from manipulating.

cs.GT↗

arXiv2025

Choosing What Game to Play without Selecting Equilibria: Inferring Safe (Pareto) Improvements in Binary Constraint Structures

We consider a setting in which a principal gets to choose which game from some given set is played by a group of agents. The principal would like to choose a game that favors one of the players, the social preferences of the players, or the principal's own preferences. Unfortunately, given the potential multiplicity of equilibria, it is conceptually unclear how to tell which of even any two games is better. Oesterheld et al. (2022) propose that we use assumptions about outcome correspondence -- i.e., about how the outcomes of different games relate -- to allow comparisons in some cases. For example, it seems reasonable to assume that isomorphic games are played isomorphically. From such assumptions we can sometimes deduce that the outcome of one game G' is guaranteed to be better than the outcome of another game G, even if we do not have beliefs about how each of G and G' will be played individually. Following Oesterheld et al., we then call G' a safe improvement on G. In this paper, we study how to derive safe improvement relations. We first show that if we are given a set of games and arbitrary assumptions about outcome correspondence between these games, deriving safe improvement relations is co-NP-complete. We then study the (in)completeness of a natural set of inference rules for outcome correspondence. We show that in general the inference rules are incomplete. However, we also show that under natural, generally applicable assumptions about outcome correspondence the rules are complete.

cs.GT↗

arXiv2025

Maximizing Social Welfare with Side Payments

We examine normal-form games in which players may \emph{pre-commit} to outcome-contingent transfers before choosing their actions. In the one-shot version of this model, Jackson and Wilkie showed that side contracting can backfire: even a game with a Pareto-optimal Nash equilibrium can devolve into inefficient equilibria once unbounded, simultaneous commitments are allowed. The root cause is a prisoner's dilemma effect, where each player can exploit her commitment power to reshape the equilibrium in her favor, harming overall welfare. To circumvent this problem we introduce a \emph{staged-commitment} protocol. Players may pledge transfers only in small, capped increments over multiple rounds, and the phase continues only with unanimous consent. We prove that, starting from any finite game $Γ$ with a non-degenerate Nash equilibrium $\vecσ$, this protocol implements every welfare-maximizing payoff profile that \emph{strictly} Pareto-improves $\vecσ$. Thus, gradual and bounded commitments restore the full efficiency potential of side payments while avoiding the inefficiencies identified by Jackson and Wilkie.

cs.GT↗

arXiv2025

Observation Interference in Partially Observable Assistance Games

We study partially observable assistance games (POAGs), a model of the human-AI value alignment problem which allows the human and the AI assistant to have partial observations. Motivated by concerns of AI deception, we study a qualitatively new phenomenon made possible by partial observability: would an AI assistant ever have an incentive to interfere with the human's observations? First, we prove that sometimes an optimal assistant must take observation-interfering actions, even when the human is playing optimally, and even when there are otherwise-equivalent actions available that do not interfere with observations. Though this result seems to contradict the classic theorem from single-agent decision making that the value of information is nonnegative, we resolve this seeming contradiction by developing a notion of interference defined on entire policies. This can be viewed as an extension of the classic result that the value of information is nonnegative into the cooperative multiagent setting. Second, we prove that if the human is simply making decisions based on their immediate outcomes, the assistant might need to interfere with observations as a way to query the human's preferences. We show that this incentive for interference goes away if the human is playing optimally, or if we introduce a communication channel for the human to communicate their preferences to the assistant. Third, we show that if the human acts according to the Boltzmann model of irrationality, this can create an incentive for the assistant to interfere with observations. Finally, we use an experimental model to analyze tradeoffs faced by the AI assistant in practice when considering whether or not to take observation-interfering actions.

cs.AI↗

arXiv2025

A dataset of questions on decision-theoretic reasoning in Newcomb-like problems

We introduce a dataset of natural-language questions in the decision theory of so-called Newcomb-like problems. Newcomb-like problems include, for instance, decision problems in which an agent interacts with a similar other agent, and thus has to reason about the fact that the other agent will likely reason in similar ways. Evaluating LLM reasoning about Newcomb-like problems is important because interactions between foundation-model-based agents will often be Newcomb-like. Some ways of reasoning about Newcomb-like problems may allow for greater cooperation between models. Our dataset contains both capabilities questions (i.e., questions with a unique, uncontroversially correct answer) and attitude questions (i.e., questions about which decision theorists would disagree). We use our dataset for an investigation of decision-theoretical capabilities and expressed attitudes and their interplay in existing models (different models by OpenAI, Anthropic, Meta, GDM, Reka, etc.), as well as models under simple prompt-based interventions. We find, among other things, that attitudes vary significantly between existing models; that high capabilities are associated with attitudes more favorable toward so-called evidential decision theory; and that attitudes are consistent across different types of questions.

cs.CL↗

arXiv2025

Promises Made, Promises Kept: Safe Pareto Improvements via Ex Post Verifiable Commitments

A safe Pareto improvement (SPI) [41] is a modification of a game that leaves all players better off with certainty. SPIs are typically proven under qualitative assumptions about the way different games are played. For example, we assume that strictly dominated strategies can be iteratively removed and that isomorphic games are played isomorphically. In this work, we study SPIs achieved through three types of ex post verifiable commitments -- promises about player behavior from which deviations can be detected by observing the game. First, we consider disarmament -- commitments not to play certain actions. Next, we consider SPIs based on token games. A token game is a game played by simply announcing an action (via cheap talk). As such, its outcome is intrinsically meaningless. However, we assume the players commit in advance to play specific (pure or correlated) strategy profiles in the original game as a function of the token game outcome. Under such commitments, the token game becomes a new, meaningful normal-form game. Finally, we consider default-conditional commitment: SPIs in settings where the players' default ways of playing the original game can be credibly revealed and hence the players can commit to act as a function of this default. We characterize the complexity of deciding whether SPIs exist in all three settings, giving a mixture of characterizations and efficient algorithms and NP- and Graph Isomorphism-hardness results.

cs.GT↗

arXiv2025

Computing Game Symmetries and Equilibria That Respect Them

Strategic interactions can be represented more concisely, and analyzed and solved more efficiently, if we are aware of the symmetries within the multiagent system. Symmetries also have conceptual implications, for example for equilibrium selection. We study the computational complexity of identifying and using symmetries. Using the classical framework of normal-form games, we consider game symmetries that can be across some or all players and/or actions. We find a strong connection between game symmetries and graph automorphisms, yielding graph automorphism and graph isomorphism completeness results for characterizing the symmetries present in a game. On the other hand, we also show that the problem becomes polynomial-time solvable when we restrict the consideration of actions in one of two ways. Next, we investigate when exactly game symmetries can be successfully leveraged for Nash equilibrium computation. We show that finding a Nash equilibrium that respects a given set of symmetries is PPAD- and CLS-complete in general-sum and team games respectively -- that is, exactly as hard as Brouwer fixed point and gradient descent problems. Finally, we present polynomial-time methods for the special cases where we are aware of a vast number of symmetries, or where the game is two-player zero-sum and we do not even know the symmetries.

cs.GT↗

arXiv2025

Multi-Agent Risks from Advanced AI

The rapid development of advanced AI agents and the imminent deployment of many instances of these agents will give rise to multi-agent systems of unprecedented complexity. These systems pose novel and under-explored risks. In this report, we provide a structured taxonomy of these risks by identifying three key failure modes (miscoordination, conflict, and collusion) based on agents' incentives, as well as seven key risk factors (information asymmetries, network effects, selection pressures, destabilising dynamics, commitment problems, emergent agency, and multi-agent security) that can underpin them. We highlight several important instances of each risk, as well as promising directions to help mitigate them. By anchoring our analysis in a range of real-world examples and experimental evidence, we illustrate the distinct challenges posed by multi-agent systems and their implications for the safety, governance, and ethics of advanced AI.

cs.MA↗

arXiv2025

Characterising Simulation-Based Program Equilibria

In Tennenholtz's program equilibrium, players of a game submit programs to play on their behalf. Each program receives the other programs' source code and outputs an action. This can model interactions involving AI agents, mutually transparent institutions, or commitments. Tennenholtz (2004) proves a folk theorem for program games, but the equilibria constructed are very brittle. We therefore consider simulation-based programs -- i.e., programs that work by running opponents' programs. These are relatively robust (in particular, two programs that act the same are treated the same) and are more practical than proof-based approaches. Oesterheld's (2019) $ε$Grounded$π$Bot is such an approach. Unfortunately, it is not generally applicable to games of three or more players, and only allows for a limited range of equilibria in two player games. In this paper, we propose a generalisation to Oesterheld's (2019) $ε$Grounded$π$Bot. We prove a folk theorem for our programs in a setting with access to a shared source of randomness. We then characterise their equilibria in a setting without shared randomness. Both with and without shared randomness, we achieve a much wider range of equilibria than Oesterheld's (2019) $ε$Grounded$π$Bot. Finally, we explore the limits of simulation-based program equilibrium, showing that the Tennenholtz folk theorem cannot be attained by simulation-based programs without access to shared randomness.

cs.GT↗

arXiv2024

Can CDT rationalise the ex ante optimal policy via modified anthropics?

In Newcomb's problem, causal decision theory (CDT) recommends two-boxing and thus comes apart from evidential decision theory (EDT) and ex ante policy optimisation (which prescribe one-boxing). However, in Newcomb's problem, you should perhaps believe that with some probability you are in a simulation run by the predictor to determine whether to put a million dollars into the opaque box. If so, then causal decision theory might recommend one-boxing in order to cause the predictor to fill the opaque box. In this paper, we study generalisations of this approach. That is, we consider general Newcomblike problems and try to form reasonable self-locating beliefs under which CDT's recommendations align with an EDT-like notion of ex ante policy optimisation. We consider approaches in which we model the world as running simulations of the agent, and an approach not based on such models (which we call 'Generalised Generalised Thirding', or GGT). For each approach, we characterise the resulting CDT policies, and prove that under certain conditions, these include the ex ante optimal policies.

cs.AI↗

arXiv2024

Computing Optimal Commitments to Strategies and Outcome-Conditional Utility Transfers

Prior work has studied the computational complexity of computing optimal strategies to commit to in Stackelberg or leadership games, where a leader commits to a strategy which is observed by one or more followers. We extend this setting to one where the leader can additionally commit to outcome-conditional utility transfers. We characterize the computational complexity of finding optimal strategies in normal-form and Bayesian games, giving a mix of efficient algorithms and NP-hardness results. Finally, we allow the leader to also commit to a signaling scheme which induces a correlated equilibrium. In this setting, optimal commitments can be found in polynomial time for arbitrarily many players.

cs.GT↗

arXiv2024

Imperfect-Recall Games: Equilibrium Concepts and Their Complexity

We investigate optimal decision making under imperfect recall, that is, when an agent forgets information it once held before. An example is the absentminded driver game, as well as team games in which the members have limited communication capabilities. In the framework of extensive-form games with imperfect recall, we analyze the computational complexities of finding equilibria in multiplayer settings across three different solution concepts: Nash, multiselves based on evidential decision theory (EDT), and multiselves based on causal decision theory (CDT). We are interested in both exact and approximate solution computation. As special cases, we consider (1) single-player games, (2) two-player zero-sum games and relationships to maximin values, and (3) games without exogenous stochasticity (chance nodes). We relate these problems to the complexity classes P, PPAD, PLS, $Σ_2^P$ , $\exists$R, and $\exists \forall$R.

cs.GT↗

arXiv2024

Game Theory with Simulation of Other Players

Game-theoretic interactions with AI agents could differ from traditional human-human interactions in various ways. One such difference is that it may be possible to simulate an AI agent (for example because its source code is known), which allows others to accurately predict the agent's actions. This could lower the bar for trust and cooperation. In this paper, we formalize games in which one player can simulate another at a cost. We first derive some basic properties of such games and then prove a number of results for them, including: (1) introducing simulation into generic-payoff normal-form games makes them easier to solve; (2) if the only obstacle to cooperation is a lack of trust in the possibly-simulated agent, simulation enables equilibria that improve the outcome for both agents; and however (3) there are settings where introducing simulation results in strictly worse outcomes for both players.

cs.GT↗

arXiv2024

Recursive Joint Simulation in Games

Game-theoretic dynamics between AI agents could differ from traditional human-human interactions in various ways. One such difference is that it may be possible to accurately simulate an AI agent, for example because its source code is known. Such an agent would then be fundamentally uncertain whether it is in the real world or in a simulation. Our aim is to explore ways of leveraging this possibility to achieve more cooperative outcomes in strategic settings. In this paper, we study an interaction between AI agents where the agents run a recursive joint simulation. That is, the agents first jointly observe a simulation of the situation they face. This simulation in turn recursively includes additional simulations (with a small chance of failure, to avoid infinite recursion), and the results of all these nested simulations are observed before an action is chosen. We show that the resulting interaction is strategically equivalent to an infinitely repeated version of the original game, allowing a direct transfer of existing results such as the various folk theorems. As evidence that the equivalence is robust, we show that it holds even when we relax some of the assumptions and that it also holds ``from the inside'' -- meaning, for an agent that finds itself inside the game and has self-locating uncertainty.

cs.AI↗

arXiv2023

Similarity-based cooperative equilibrium

As machine learning agents act more autonomously in the world, they will increasingly interact with each other. Unfortunately, in many social dilemmas like the one-shot Prisoner's Dilemma, standard game theory predicts that ML agents will fail to cooperate with each other. Prior work has shown that one way to enable cooperative outcomes in the one-shot Prisoner's Dilemma is to make the agents mutually transparent to each other, i.e., to allow them to access one another's source code (Rubinstein 1998, Tennenholtz 2004) -- or weights in the case of ML agents. However, full transparency is often unrealistic, whereas partial transparency is commonplace. Moreover, it is challenging for agents to learn their way to cooperation in the full transparency setting. In this paper, we introduce a more realistic setting in which agents only observe a single number indicating how similar they are to each other. We prove that this allows for the same set of cooperative outcomes as the full transparency setting. We also demonstrate experimentally that cooperation can be learned using simple ML methods.

cs.GT↗

arXiv2023

A New Formalism, Method and Open Issues for Zero-Shot Coordination

In many coordination problems, independently reasoning humans are able to discover mutually compatible policies. In contrast, independently trained self-play policies are often mutually incompatible. Zero-shot coordination (ZSC) has recently been proposed as a new frontier in multi-agent reinforcement learning to address this fundamental issue. Prior work approaches the ZSC problem by assuming players can agree on a shared learning algorithm but not on labels for actions and observations, and proposes other-play as an optimal solution. However, until now, this "label-free" problem has only been informally defined. We formalize this setting as the label-free coordination (LFC) problem by defining the label-free coordination game. We show that other-play is not an optimal solution to the LFC problem as it fails to consistently break ties between incompatible maximizers of the other-play objective. We introduce an extension of the algorithm, other-play with tie-breaking, and prove that it is optimal in the LFC problem and an equilibrium in the LFC game. Since arbitrary tie-breaking is precisely what the ZSC setting aims to prevent, we conclude that the LFC problem does not reflect the aims of ZSC. To address this, we introduce an alternative informal operationalization of ZSC as a starting point for future work.

cs.AI↗