arXiv subjects

Haifeng Xu

Publications and source records attributed to Haifeng Xu.

At least 37 records · Page 2Linked to original sources

arXiv2026

Partitionable Diffractive Neural Networks for Multifunctional Optical Operations

Diffractive neural network (DNN), which can perform machine learning tasks based on the light propagation and diffraction, has recently emerged as a promising optical computing paradigm due to its high parallel processing speed and low power consumption nature. However, existing diffractive network architectures face challenges in implementing functional reconfiguration. Once a diffractive neural network is fabricated, its functionality is fixed. Deploying such systems for different tasks typically requires reconstructing the entire physical setup, which significantly compromises hardware efficiency in practical applications. In this work, we propose the multifunctional partitionable diffractive neural networks (PDNNs) that can generate networks with additional capabilities by stacking multiple sub-modules with independent functions in the horizontal direction. Each submodule functions as an independent diffractive network capable of performing specific imaging or classification tasks. When these submodules are combined, they can form a new network with additional functionalities. Moreover, assembling these submodules in different configurations enables structures with diverse functions. This powerful PDNN framework demonstrates remarkable advantages in flexibility and reconfigurability for multitask operations, opening a new pathway for realizing multifunctional and integrated optical artificial intelligence systems.

physics.optics↗

arXiv2026

An AI-guided mechanotyping instrument for fully automated oocyte quality assessment

The mechanical properties of oocytes are regarded as important indicators of their developmental potential. During fertilization, deviations from the normal mechanical range can hinder sperm penetration, ultimately reducing fertilization efficiency and compromising embryo quality. However, current methods for measuring oocyte mechanics often suffer from serious cellular damage, low automation levels, and large measurement errors. To address these limitations, we developed an AI-guided micronewton-scale mechanical measurement system for safe and automated oocyte quality assessment. The system integrates voice interaction with automated experimental workflows to control a magnetically actuated microgripper, which applies defined loading forces to induce micron-scale compressive deformation of the oocyte. Combined with AI-assisted object detection and image segmentation algorithms, the system captures cellular deformation in real time, enabling precise calculation of the oocyte's compressive modulus. This measurement system enables automated, quantitative, and non-destructive evaluation of oocyte mechanical properties, providing an effective approach for oocyte quality screening in in vitro fertilization (IVF) and other assisted reproductive technologies (ART).

physics.bio-ph↗

arXiv2025

Data-Driven Mechanism Design: Jointly Eliciting Preferences and Information

We study mechanism design in environments where agents have private preferences and private information about a common payoff-relevant state. In such settings with multi-dimensional types, standard mechanisms fail to implement efficient allocations. We address this limitation by proposing data-driven mechanisms that condition transfers on additional post-allocation information, modeled as an estimator of the payoff-relevant state. Our mechanisms extend the classic Vickrey-Clarke-Groves framework. We show they achieve exact implementation in posterior equilibrium when the state is fully revealed or utilities are affine in an unbiased estimator. With a consistent estimator, they achieve approximate implementation that converges to exact implementation as the estimator converges, and we provide bounds on the convergence rate. We demonstrate applications to digital advertising auctions and AI shopping assistants, where user engagement naturally reveals relevant information, and to procurement auctions with consumer spot markets, where additional information arises from a pricing game played by the same agents.

econ.TH↗

arXiv2025

LLM-as-a-Prophet: Understanding Predictive Intelligence with Prophet Arena

Forecasting is not only a fundamental intellectual pursuit but also is of significant importance to societal systems such as finance and economics. With the rapid advances of large language models (LLMs) trained on Internet-scale data, it raises the promise of employing LLMs to forecast real-world future events, an emerging paradigm we call "LLM-as-a-Prophet". This paper systematically investigates such predictive intelligence of LLMs. To this end, we build Prophet Arena, a general evaluation benchmark that continuously collects live forecasting tasks and decomposes each task into distinct pipeline stages, in order to support our controlled and large-scale experimentation. Our comprehensive evaluation reveals that many LLMs already exhibit impressive forecasting capabilities, reflected in, e.g., their small calibration errors, consistent prediction confidence and promising market returns. However, we also uncover key bottlenecks towards achieving superior predictive intelligence via LLM-as-a-Prophet, such as LLMs' inaccurate event recalls, misunderstanding of data sources and slower information aggregation compared to markets when resolution nears.

cs.AI↗

arXiv2025

The Publication Choice Problem

Researchers strategically choose where to submit their work in order to maximize its impact, and these publication decisions in turn determine venues' impact factors. To analyze how individual publication choices both respond to and shape venue impact, we introduce a game-theoretic framework, coined the Publication Choice Problem, that captures this two-way interplay. We show the existence of a pure-strategy equilibrium in the Publication Choice Problem and its uniqueness under binary researcher types. Our characterizations of the equilibrium properties offer insights about what publication behaviors better indicate a researcher's impact level. Through equilibrium analysis, we further investigate how labeling papers with ``spotlight'' affects the impact factor of venues in the research community. Our analysis shows that competitive venue labeling top papers with ``spotlight'' may decrease the overall impact of other venues in the community, while less competitive venues with ``spotlight'' labeling have the opposite impact.

cs.GT↗

arXiv2025

Data Auctions for Retrieval Augmented Generation

We study the problem of data selling for Retrieval Augmented Generation (RAG) tasks in Generative AI applications. We model each buyer's valuation of a dataset with a natural coverage-based valuation function that increases with the inclusion of more relevant data points that would enhance responses to anticipated queries. Motivated by issues such as data control and prior-free revenue maximization, we focus on the scenario where each data point can be allocated to only one buyer. We show that the problem of welfare maximization in this setting is NP-hard even with two bidders, but design a polynomial-time $(1-1/e)$ approximation algorithm for any number of bidders. Unfortunately, however, this efficient allocation algorithm fails to be incentive compatible. The crux of our approach is a carefully tailored post-processing step called data burning which retains the $(1-1/e)$ approximation factor but achieves incentive compatibility. Our thorough experiments on synthetic and real-world image and text datasets demonstrate the practical effectiveness of our algorithm compared to popular baseline algorithms for combinatorial auctions.

cs.GT↗

arXiv2025

Accelerated Distance-adaptive Methods for Hölder Smooth and Convex Optimization

This paper introduces new parameter-free first-order methods for convex optimization problems in which the objective function exhibits Hölder smoothness. Inspired by the recently proposed distance-over-gradient (DOG) technique, we propose an accelerated distance-adaptive method which achieves optimal anytime convergence rates for Hölder smooth problems without requiring prior knowledge of smoothness parameters or explicit parameter tuning. Importantly, our parameter-free approach removes the necessity of specifying target accuracy in advance, addressing a limitation found in the universal fast gradient methods (Nesterov, Yu. \textit{Mathematical Programming}, 2015). For convex stochastic optimization, we further present a parameter-free accelerated method that eliminates the need for line-search procedures. Preliminary experimental results highlight the effectiveness of our approach on convex nonsmooth problems and its advantages over existing parameter-free or accelerated methods.

math.OC↗

arXiv2025

Learning Personalized Ad Impact via Contextual Reinforcement Learning under Delayed Rewards

Online advertising platforms use automated auctions to connect advertisers with potential customers, requiring effective bidding strategies to maximize profits. Accurate ad impact estimation requires considering three key factors: delayed and long-term effects, cumulative ad impacts such as reinforcement or fatigue, and customer heterogeneity. However, these effects are often not jointly addressed in previous studies. To capture these factors, we model ad bidding as a Contextual Markov Decision Process (CMDP) with delayed Poisson rewards. For efficient estimation, we propose a two-stage maximum likelihood estimator combined with data-splitting strategies, ensuring controlled estimation error based on the first-stage estimator's (in)accuracy. Building on this, we design a reinforcement learning algorithm to derive efficient personalized bidding strategies. This approach achieves a near-optimal regret bound of $\tilde{O}{(dH^2\sqrt{T})}$, where $d$ is the contextual dimension, $H$ is the number of rounds, and $T$ is the number of customers. Our theoretical findings are validated by simulation experiments.

cs.LG↗

arXiv2025

Robust Optimization with Diffusion Models for Green Security

In green security, defenders must forecast adversarial behavior, such as poaching, illegal logging, and illegal fishing, to plan effective patrols. These behavior are often highly uncertain and complex. Prior work has leveraged game theory to design robust patrol strategies to handle uncertainty, but existing adversarial behavior models primarily rely on Gaussian processes or linear models, which lack the expressiveness needed to capture intricate behavioral patterns. To address this limitation, we propose a conditional diffusion model for adversary behavior modeling, leveraging its strong distribution-fitting capabilities. To the best of our knowledge, this is the first application of diffusion models in the green security domain. Integrating diffusion models into game-theoretic optimization, however, presents new challenges, including a constrained mixed strategy space and the need to sample from an unnormalized distribution to estimate utilities. To tackle these challenges, we introduce a mixed strategy of mixed strategies and employ a twisted Sequential Monte Carlo (SMC) sampler for accurate sampling. Theoretically, our algorithm is guaranteed to converge to an epsilon equilibrium with high probability using a finite number of iterations and samples. Empirically, we evaluate our approach on both synthetic and real-world poaching datasets, demonstrating its effectiveness.

cs.CY↗

arXiv2025

Strategic Filtering for Content Moderation: Free Speech or Free of Distortion?

User-generated content (UGC) on social media platforms is vulnerable to incitements and manipulations, necessitating effective regulations. To address these challenges, those platforms often deploy automated content moderators tasked with evaluating the harmfulness of UGC and filtering out content that violates established guidelines. However, such moderation inevitably gives rise to strategic responses from users, who strive to express themselves within the confines of guidelines. Such phenomena call for a careful balance between: 1. ensuring freedom of speech -- by minimizing the restriction of expression; and 2. reducing social distortion -- measured by the total amount of content manipulation. We tackle the problem of optimizing this balance through the lens of mechanism design, aiming at optimizing the trade-off between minimizing social distortion and maximizing free speech. Although determining the optimal trade-off is NP-hard, we propose practical methods to approximate the optimal solution. Additionally, we provide generalization guarantees determining the amount of finite offline data required to approximate the optimal moderator effectively.

cs.LG↗

arXiv2025

Metasurface-empowered freely-arrangeable multi-task diffractive neural networks with weighted training

Recent advancements in optical computing have garnered considerable research interests owing to its ener-gy-efficient operation and ultralow latency characteristics. As an emerging framework in this domain, dif-fractive deep neural networks (D2NNs) integrate deep learning algorithms with optical diffraction principles to perform computational tasks at light speed without requiring additional energy consumption. Neverthe-less, conventional D2NN architectures face functional limitations and are typically constrained to single-task operations or necessitating additional costs and structures for functional reconfiguration. Here, an arrangea-ble diffractive neural network (A-DNN) that achieves low-cost reconfiguration and high operational versa-tility by means of diffractive layer rearrangement is presented. Our architecture enables dynamic reordering of pre-trained diffractive layers to accommodate diverse computational tasks. Additionally, we implement a weighted multi-task loss function that allows precise adjustment of task-specific performances. The efficacy of the system is demonstrated by both numerical simulations and experimental validations of recognizing handwritten digits and fashions at terahertz frequencies. Our proposed architecture can greatly expand the flexibility of D2NNs at a low cost, providing a new approach for realizing high-speed, energy-efficient ver-satile artificial intelligence systems.

physics.app-ph↗

arXiv2025

Tokenized Bandit for LLM Decoding and Alignment

We introduce the tokenized linear bandit (TLB) and multi-armed bandit (TMAB), variants of linear and stochastic multi-armed bandit problems inspired by LLM decoding and alignment. In these problems, at each round $t \in [T]$, a user submits a query (context), and the decision maker (DM) sequentially selects a token irrevocably from a token set. Once the sequence is complete, the DM observes a random utility from the user, whose expectation is presented by a sequence function mapping the chosen token sequence to a nonnegative real value that depends on the query. In both problems, we first show that learning is impossible without any structure on the sequence function. We introduce a natural assumption, diminishing distance with more commons (DDMC), and propose algorithms with regret $\tilde{O}(L\sqrt{T})$ and $\tilde{O}(L\sqrt{T^{2/3}})$ for TLB and TMAB, respectively. As a side product, we obtain an (almost) optimality of the greedy decoding for LLM decoding algorithm under DDMC, which justifies the unresaonable effectiveness of greedy decoding in several tasks. This also has an immediate application to decoding-time LLM alignment, when the misaligned utility can be represented as the frozen LLM's utility and a linearly realizable latent function. We finally validate our algorithm's performance empirically as well as verify our assumptions using synthetic and real-world datasets.

cs.LG↗

arXiv2025

Robust Stackelberg Equilibria

This paper provides a systematic study of the robust Stackelberg equilibrium (RSE), which naturally generalizes the widely adopted solution concept of the strong Stackelberg equilibrium (SSE). The RSE accounts for any possible up-to-$δ$ suboptimal follower responses in Stackelberg games and is adopted to improve the robustness of the leader's strategy. While a few variants of robust Stackelberg equilibrium have been considered in previous literature, the RSE solution concept we consider is importantly different -- in some sense, it relaxes previously studied robust Stackelberg strategies and is applicable to much broader sources of uncertainties. We provide a thorough investigation of several fundamental properties of RSE, including its utility guarantees, algorithmics, and learnability. We first show that the RSE we defined always exists and thus is well-defined. Then we characterize how the leader's utility in RSE changes with the robustness level considered. On the algorithmic side, we show that, in sharp contrast to the tractability of computing an SSE, it is NP-hard to obtain a fully polynomial approximation scheme (FPTAS) for any constant robustness level. Nevertheless, we develop a quasi-polynomial approximation scheme (QPTAS) for RSE. Finally, we examine the learnability of the RSE in a natural learning scenario, where both players' utilities are not known in advance, and provide almost tight sample complexity results on learning the RSE. As a corollary of this result, we also obtain an algorithm for learning SSE, which strictly improves a key result of Bai et al. [2021] in terms of both utility guarantee and computational efficiency.

cs.GT↗

arXiv2025

Optimal Pricing for Data-Augmented AutoML Marketplaces

Organizations often lack sufficient data to effectively train machine learning (ML) models, while others possess valuable data that remains underutilized. Data markets promise to unlock substantial value by matching data suppliers with demand from ML consumers. However, market design involves addressing intricate challenges, including data pricing, fairness, robustness, and strategic behavior. In this paper, we propose a pragmatic data-augmented AutoML market that seamlessly integrates with existing cloud-based AutoML platforms such as Google's Vertex AI and Amazon's SageMaker. Unlike standard AutoML solutions, our design automatically augments buyer-submitted training data with valuable external datasets, pricing the resulting models based on their measurable performance improvements rather than computational costs as the status quo. Our key innovation is a pricing mechanism grounded in the instrumental value - the marginal model quality improvement - of externally sourced data. This approach bypasses direct dataset pricing complexities, mitigates strategic buyer behavior, and accommodates diverse buyer valuations through menu-based options. By integrating automated data and model discovery, our solution not only enhances ML outcomes but also establishes an economically sustainable framework for monetizing external data.

cs.LG↗

arXiv2025

Cer-Eval: Certifiable and Cost-Efficient Evaluation Framework for LLMs

As foundation models continue to scale, the size of trained models grows exponentially, presenting significant challenges for their evaluation. Current evaluation practices involve curating increasingly large datasets to assess the performance of large language models (LLMs). However, there is a lack of systematic analysis and guidance on determining the sufficiency of test data or selecting informative samples for evaluation. This paper introduces a certifiable and cost-efficient evaluation framework for LLMs. Our framework adapts to different evaluation objectives and outputs confidence intervals that contain true values with high probability. We use ``test sample complexity'' to quantify the number of test points needed for a certifiable evaluation and derive tight bounds on test sample complexity. Based on the developed theory, we develop a partition-based algorithm, named Cer-Eval, that adaptively selects test points to minimize the cost of LLM evaluation. Real-world experiments demonstrate that Cer-Eval can save 20% to 40% test points across various benchmarks, while maintaining an estimation error level comparable to the current evaluation process and providing a 95% confidence guarantee.

stat.ML↗

arXiv2025

Is Learning Effective in Dynamic Strategic Interactions? Evidence from Stackelberg Games

In many settings of interest, a policy is set by one party, the leader, in order to influence the action of another party, the follower, where the follower's response is determined by some private information. A natural question to ask is, can the leader improve their strategy by learning about the unknown follower through repeated interactions? A well known folk theorem from dynamic pricing, a special case of this leader-follower setting, would suggest that the leader cannot learn effectively from the follower when the follower is fully strategic, leading to a large literature on learning in strategic settings that relies on limiting the strategic space of the follower in order to provide positive results. In this paper, we study dynamic Bayesian Stackelberg games, where a leader and a \emph{fully strategic} follower interact repeatedly, with the follower's type unknown. Contrary to existing results, we show that the leader can improve their utility through learning in repeated play. Using a novel average-case analysis, we demonstrate that learning is effective in these settings, without needing to weaken the follower's strategic space. Importantly, this improvement is not solely due to the leader's ability to commit, nor does learning simply substitute for communication between the parties. We provide an algorithm, based on a mixed-integer linear program, to compute the optimal leader policy in these games and develop heuristic algorithms to approximate the optimal dynamic policy more efficiently. Through simulations, we compare the efficiency and runtime of these algorithms against static policies.

cs.GT↗

arXiv2025

How to Strategize Human Content Creation in the Era of GenAI?

Generative AI (GenAI) will have significant impact on content creation platforms. In this paper, we study the dynamic competition between a GenAI and a human contributor. Unlike the human, the GenAI's content only improves when more contents are created by the human over time; however, GenAI has the advantage of generating content at a lower cost. We study the algorithmic problem in this dynamic competition model about how the human contributor can maximize her utility when competing against the GenAI for content generation over a set of topics. In time-sensitive content domains (e.g., news or pop music creation) where contents' value diminishes over time, we show that there is no polynomial time algorithm for finding the human's optimal (dynamic) strategy, unless the randomized exponential time hypothesis is false. Fortunately, we are able to design a polynomial time algorithm that naturally cycles between myopically optimizing over a short time window and pausing and provably guarantees an approximation ratio of $\frac{1}{2}$. We then turn to time-insensitive content domains where contents do not lose their value (e.g., contents on history facts). Interestingly, we show that this setting permits a polynomial time algorithm that maximizes the human's utility in the long run. Finally, we conduct simulations that demonstrate the advantage of our algorithms in comparison to a collection of baselines.

cs.GT↗

arXiv2025

Competitive Information Design for Pandora's Box

We study a natural competitive-information-design strategic variant for the celebrated Pandora's Box problem (Weitzman, 1979), where each box is associated with a strategic information sender who can design what information about the box's prize value to be revealed to the agent when the agent inspects the box. This variant with strategic boxes is motivated by a wide range of real-world economic applications for Pandora's Box. Our contributions are three-fold: (1) given the boxes' information policies, we characterize the agent's optimal search and stopping strategy; (2) we fully characterize the pure symmetric equilibrium for the game of boxes' competitive information revelation in a symmetric environment; and (3) we reveal various insights regarding information competition and the resultant agent payoff at equilibrium, and additionally, we study informational properties of Pandora's Box by establishing an intrinsic connection between informativeness of any box's value distribution and the utility order of the search agent.

cs.GT↗