arXiv subjects

Zhengye Han

Publications and source records attributed to Zhengye Han.

9 recordsLinked to original sources

arXiv2026

From Emergence to Reversibility: Toward Reversible Multi-Agent Systems

Multi-agent systems become capable through institutional mechanisms, such as rules that route information to agents that act on it. These mechanisms are also natural places to restrict the system. Yet as agents learn, share memory, delegate, and repair failures, the collective can find other paths to the same effects: a restriction on the original mechanism still operates but no longer stops them, while routine performance and compliance checks pass. We propose \textbf{Reversible Multi-agent Systems} as a research agenda: collectives that grow while an authorized supervisor can still restrict specified effects, keep them restricted as agents adapt, and reconstitute useful cooperation under renewed authority. What is reversed is the ability to produce a specified effect; restrictions must hold even if learned knowledge persists. The goal is to preserve effective intervention, not a particular mechanism: at every step of growth, some authorized intervention must keep a specified prohibited effect rare while maintaining required service. We identify three research problems: tracking what capabilities depend on, handing off control before an old mechanism loses its effect, and separating legitimate repair from unauthorized reconstruction. We outline how to evaluate control across growth paths, back control claims with evidence, and allocate the rights to restrict and restore.

cs.MA↗

arXiv2026

Where Do Multi-Agent Systems Fail? Evidence-Grounded Diagnosis of Collective Mechanisms

When a multi-agent system answers correctly, it is tempting to conclude that its agents shared, checked, and used information as intended. Yet a system can break one of its collective mechanisms, the rules that govern how agents route, admit, store, and act on shared information, and still return the right answer, while a wrong answer rarely reveals which mechanism failed. We ask what evidence from an execution is sufficient to conclude that a particular mechanism was violated. Our answer is a diagnostic contract, which separates what counts as a violation from which execution records can establish one, and concludes that a violation is supported, ruled out, or unknown; removing records can make this conclusion unknown but never reverse it. We test contracts for four mechanisms by replaying executions from the step where a mechanism acts, once unchanged, once with the mechanism broken, and once with it restored. Broken mechanisms often left the answer correct. An LLM diagnoser detected many more violations from internal records than from public outputs, yet with identical records a generic prompt often claimed certainty the records did not support, which prompts stating the contracts largely avoided. The contracts also applied, in narrow form, to mechanisms in independently developed systems, but a diagnostic behavior that was nearly perfect on our benchmark degraded on an independently developed workflow. A correct outcome is therefore no substitute for records of how collective mechanisms operated, and agreement on one benchmark does not show that a diagnoser transfers to another system.

cs.MA↗

arXiv2026

When Do Institutions Beat Intelligence?

More capable agents do not necessarily form a more capable collective. A multi-agent system may jointly possess sufficient information yet fail because evidence is poorly routed, unreliable reports enter public belief, correlated claims masquerade as independent support, shared state becomes stale or strategically distorted, or useful evidence is exposed through an ineffective action interface. We ask when additional resources should improve the reasoner and when they should instead change the institutional structure through which the collective forms and acts on public information. Drawing on functional distinctions from research on group decision making and distributed cognition, we construct controlled artificial ecologies around four loci of collective failure: access and routing, admission and dependence, state maintenance and incentives, and representation and action. Across these ecologies, we separately vary model capability and institutional structure, pairing positive interventions with matched reasoning baselines and mechanism-breaking controls. The experiments reveal a consistent boundary: institutions help when they repair failures in how a collective constructs usable public state, but lose their advantage when their signals are uninformative or uncheckable, when stronger intelligence can perform the same transformation directly, or when the resulting state cannot support reliable action. Our results recast the choice between intelligence and institutions as a diagnosis of where collective reasoning fails.

cs.MA↗

arXiv2026

RCSP: Risk-Sensitive Conjectural Scenario Planning for Safe Dynamic Robot Navigation

Mobile robots can fail before they collide: a velocity that is safe now may commit the robot to a passage that moving obstacles will soon close. We study this predictive near-miss commitment problem and propose Risk-Sensitive Conjectural Scenario Planning (RCSP), a planning layer that evaluates candidate commands against plausible short-horizon obstacle futures. RCSP maintains a lightweight belief over local motion conjectures, samples future interactions, penalizes high-risk tails, and executes through a local safety check. In controlled MuJoCo bottleneck tasks, the RCSP planner reaches the goal without collisions and yields higher secondary safety and path-quality point estimates than a non-adaptive predictor, with additional latency. In ROS2/Gazebo, adding the local safety layer to a standard Nav2 stack reduces dynamic near-miss failures. On official DynaBARN/Jackal transfer, tuned DWA and TEB remain stronger on strict benchmark success, revealing the boundary of the approach. These simulation results position RCSP as a predictive-risk module that complements existing navigation stacks in dynamic bottleneck regimes.

cs.RO↗

arXiv2026

Performative Scenario Optimization

This paper introduces a performative scenario optimization framework for decision-dependent chance-constrained problems. Unlike classical stochastic optimization, we account for the feedback loop where decisions actively shape the underlying data-generating process. We define performative solutions as self-consistent equilibria and establish their existence using Kakutani's fixed-point theorem. To ensure computational tractability without requiring an explicit model of the environment, we propose a model-free, scenario-based approximation that alternates between data generation and optimization. Under mild regularity conditions, we prove that a stochastic fixed-point iteration, equipped with a logarithmic sample size schedule, converges almost surely to the unique performative solution. The effectiveness of the proposed framework is demonstrated through an emerging AI safety application: deploying performative guardrails against Large Language Model (LLM) jailbreaks. Numerical results confirm the co-evolution and convergence of the guardrail classifier and the induced adversarial prompt distribution to a stable equilibrium.

cs.GT↗

arXiv2026

Split-Merge Dynamics for Shapley-Fair Coalition Formation

Coalition formation is often modeled as a static equilibrium problem, neglecting the dynamic processes governing how agents self-organize. This paper proposes a dynamic split-and-merge framework that balances two conflicting economic forces: individual fairness and collective efficiency. We introduce a control-theoretic mechanism where topological operations are driven by distinct signals: splits are triggered by fairness violations (specifically, negative Shapley values representing "agent-responsible inefficiency"), while merges are driven by strict surplus improvements (superadditivity). We prove that these dynamics converge in finite time to a specific class of steady states termed Shapley-Fair and Merge-Stable (SFMS) partitions. Convergence is established via a vector Lyapunov function tracking aggregate fairness deficits and system surplus, leveraging a discrete-time LaSalle invariance principle. Numerical case studies on a 10-player game demonstrate the algorithm's ability to resolve fairness tensions and reach stable configurations, providing a rigorous foundation for endogenous coalition formation in dynamic environments.

cs.GT↗

arXiv2026

Learning, Misspecification, and Cognitive Arbitrage in Linear-Quadratic Network Games

We study strategic interaction in linear-quadratic network games where agents act on subjective, misspecified models of their environment. Agents observe noisy aggregate signals generated by local network externalities and interpret them through simplified conjectures, such as constant or mean-field representations. We characterize the long-run behavior using the Berk-Nash equilibrium (BNE) concept, establishing conditions under which BNE diverges from the Nash equilibrium of the perfectly specified game. We quantify this divergence using a Value of Misspecification (VoM) metric. Building on this framework, we introduce "cognitive arbitrage" -- a design paradigm where a system designer strategically shapes agents' conjectures via minimal observation distortions to steer equilibrium outcomes. We formulate the cognitive arbitrage problem as a Stackelberg optimization with closed-form solutions and prove the convergence of a two-time-scale learning algorithm to the optimal BNE. Our results provide a principled framework for influencing behavior in networked systems with bounded rationality, offering a new perspective on mechanism design that operates on agents' representations rather than their incentives.

cs.GT↗

arXiv2026

A Mathematical Programming Approach to Computing and Learning Berk--Nash Equilibria in Infinite-Horizon MDPs

We study sequential decision-making when the agent's internal model class is misspecified. Within the infinite-horizon Berk-Nash framework, stable behavior arises as a fixed point: the agent acts optimally relative to a subjective model, while that model is statistically consistent with the long-run data endogenously generated by the policy itself. We provide a rigorous characterization of this equilibrium via coupled linear programs and a bilevel optimization formulation. To address the intrinsic non-smoothness of standard best-response correspondences, we introduce entropy regularization, establishing the existence of a unique soft Bellman fixed point and a smooth objective. Exploiting this regularity, we develop an online learning scheme that casts model selection as an adversarial bandit problem using an EXP3-type update, augmented by a novel conjecture-set zooming mechanism that adaptively refines the parameter space. Numerical results demonstrate effective exploration-exploitation trade-offs, convergence to the KL-minimizing model, and sublinear regret.

cs.GT↗

arXiv2026

Toward a Dynamic Stackelberg Game-Theoretic Framework for Agentic AI Defense Against LLM Jailbreaking

This paper proposes a game theoretic framework that models the interaction between prompt engineers and large language models (LLMs) as a two player extensive form game coupled with a Rapidly exploring Random Trees (RRT) search over prompt space. The attacker incrementally samples, extends, and tests prompts, while the LLM chooses to accept, reject, or redirect, leading to terminal outcomes of Safe Interaction, Blocked, or Jailbreak. Embedding RRT exploration inside the extensive form game captures both the discovery phase of jailbreak strategies and the strategic responses of the model. Furthermore, we show that the defender behavior can be interpreted through a local Stackelberg equilibrium condition, which explains when the attacker can no longer obtain profitable prompt deviations and provides a theoretical lens for understanding the effectiveness of our Purple Agent defense. The resulting game tree thus offers a principled foundation for evaluating, interpreting, and hardening LLM guardrails.

cs.AI↗