arXiv subjects

Wanying Huang

Publications and source records attributed to Wanying Huang.

7 recordsLinked to original sources

arXiv2026

A coarse geometric approach to graph layout problems

We define a range of new coarse geometric invariants based on various graph-theoretic measures of complexity for finite graphs, including: treewidth, pathwidth, cutwidth and bandwidth. We prove that, for bounded degree graphs, these invariants can be used to define functions which satisfy a strong monotonicity property, namely they are monotonically non-decreasing with respect to a large-scale geometric generalisation of graph inclusion, and as such have potential applications in coarse geometry and geometric group theory. On the graph-theoretic side, we prove asymptotically optimal bounds on most of the above widths for the family of all finite subgraphs of any bounded degree graph whose separation profile is known to be of the form $r^a\log(r)^b$ for some $a>0$. This large class includes Diestel-Leader graphs, all Cayley graphs of non-virtually cyclic polycyclic groups, uniform lattices in almost all connected unimodular Lie groups, and many hyperbolic groups.

math.MG↗

arXiv2026

Should I State or Should I Show? Aligning AI with Human Preferences

As AI agents become more autonomous, properly aligning their objectives with human preferences becomes increasingly important. We study how effectively an AI agent learns a human principal's preference in choice under risk via stated versus revealed preferences. We conduct an online experiment in which subjects state their preferences through written instructions ("prompts") and reveal them through choices in a series of binary lottery questions ("data"). We find that on average, an AI agent given revealed-preference data predicts subjects' choices more accurately than an AI agent given stated-preference prompts. Further analysis suggests that the gap is driven by subjects' difficulty in translating their own preferences into written instructions. When given a choice between which information source to give to an AI agent, a large portion of subjects fail to select the more informative one. Moreover, when predictions from the two sources conflict, we find that the AI agent aligns more frequently with the prompt, despite its lower accuracy. Overall, these results highlight the revealed preference approach as a powerful mechanism for communicating human preferences to AI agents, but its success depends on careful implementation.

econ.GN↗

arXiv2026

On the Inefficiency of Social Learning

We study whether a social planner can improve the efficiency of learning, measured by the expected total welfare loss, in a sequential decision-making environment. Agents arrive in order and each makes a binary action based on their private signal and the social information they observe. The planner can intervene by jointly designing the social information disclosed to agents and offering monetary transfers contingent on agents' actions. We show that, despite such flexibility, efficient learning cannot be restored with a finite budget: whenever learning is inefficient without intervention, no combination of information disclosure and transfers can achieve efficient learning while keeping total expected transfers finite.

econ.TH↗

arXiv2025

Correlated Perfect Equilibrium

We propose a refinement of correlated equilibrium based on mediator errors, called correlated perfect equilibrium (CPE). In finite games, the set of CPE is nonempty and forms a finite union of convex sets. Like perfect equilibrium, a CPE never assigns positive probability to any weakly dominated strategy. We provide a dual representation of CPE and demonstrate how it differs from two existing refinements of correlated equilibrium--acceptable correlated equilibrium (Myerson, 1986) and perfect direct correlated equilibrium (Dhillon-Mertens, 1996)--through examples.

econ.TH↗

arXiv2025

Learning about informativeness

We study a sequential social learning model in which there is uncertainty about the informativeness of a common signal-generating process. Rational agents arrive in order and make decisions based on the past actions of others and their private signals. We show that, in this setting, asymptotic learning about informativeness is not guaranteed and depends crucially on the relative tail distributions of the private beliefs induced by uninformative and informative signals. We identify the phenomenon of perpetual disagreement as the cause of learning and characterize learning in the canonical Gaussian environment.

econ.TH↗

arXiv2024

The Emergence of Fads in a Changing World

We study how fads emerge under social learning in a changing environment. We consider a simple sequential social learning model where rational agents arrive in order, each acting only once, and the underlying unknown state constantly evolves. Each agent receives a private signal, observes all past actions of others, and chooses an action to match the current state. Because the state changes over time, cascades cannot last forever, and actions also fluctuate. We show that despite the rise of temporary information cascades, in the long run, actions change more often than the state. This result provides a theoretical foundation for faddish behavior in which people often change their actions more frequently than necessary.

econ.TH↗

arXiv2024

Learning in Repeated Interactions on Networks

We study how long-lived, rational agents learn in a social network. In every period, after observing the past actions of his neighbors, each agent receives a private signal, and chooses an action whose payoff depends only on the state. Since equilibrium actions depend on higher order beliefs, it is difficult to characterize behavior. Nevertheless, we show that regardless of the size and shape of the network, the utility function, and the patience of the agents, the speed of learning in any equilibrium is bounded from above by a constant that only depends on the private signal distribution.

econ.TH↗