arXiv subjects

Xiaoxi Wang

Publications and source records attributed to Xiaoxi Wang.

8 recordsLinked to original sources

arXiv2026

Learning Long-Term Educational Investment Policies under Residential Sorting

Allocating public-school investment effectively and fairly is difficult when school access depends on residence. School improvements can raise nearby housing demand and prices, reshape enrollment, and potentially limit access for lower-income households. These effects evolve as residential sorting changes school composition, quality, and future investment needs. Existing approaches often study school funding, household choice, and housing markets separately, while static models can miss their interconnected, long-term effects. We address this gap with a dynamic multi-agent framework that links government investment, household sorting, housing prices, population turnover, enrollment, and evolving school quality. A government planner uses reinforcement learning (RL) to identify multiyear allocation policies that account for household responses while balancing aggregate educational access and equity. In simulations, our RL-based policy attains the highest access level (0.4780) and second-lowest access Gini coefficient (0.0164) among representative baselines, demonstrating a favorable effectiveness-equity balance. The results also indicate reduced socioeconomic stratification in educational access. By making education-housing feedback explicit, our framework supports long-term analysis of how school investment shapes educational opportunity over time.

cs.MA↗

arXiv2024

Balancing the AI Strength of Roles in Self-Play Training with Regret Matching+

When training artificial intelligence for games encompassing multiple roles, the development of a generalized model capable of controlling any character within the game presents a viable option. This strategy not only conserves computational resources and time during the training phase but also reduces resource requirements during deployment. training such a generalized model often encounters challenges related to uneven capabilities when controlling different roles. A simple method is introduced based on Regret Matching+, which facilitates a more balanced performance of strength by the model when controlling various roles.

cs.AI↗

arXiv2023

Integrated O- and C-band Silicon-Lithium Niobate Mach-Zehnder Modulators with 100 GHz Bandwidth, Low Voltage, and Low Loss

Broadband integrated thin-film lithium niobate (TFLN) electro-optic modulators (EOM) are desirable for optical communications and signal processing in both the O-band (1310 nm) and C-band (1550 nm). To address these needs, we design and demonstrate Mach-Zehnder (MZ) EOM devices in a hybrid platform based on TFLN bonded to foundry-fabricated silicon photonic waveguides. Using a single silicon lithography step and a single bonding step, we realize MZ EOM devices which cover both wavelength ranges on the same chip. The EOM devices achieve 100 GHz EO bandwidth (referenced to 1 GHz) and about 2-3 V$.$cm figure-of-merit ($V_πL$) with low on-chip optical loss in both the O-band and C-band.

physics.optics↗

arXiv2022

110 GHz, 110 mW Hybrid Silicon-Lithium Niobate Mach-Zehnder Modulator

High bandwidth, low voltage electro-optic modulators with high optical power handling capability are important for improving the performance of analog optical communications and RF photonic links. Here we designed and fabricated a thin-film lithium niobate (LN) Mach-Zehnder modulator (MZM) which can handle high optical power of 110 mW, while having 3-dB bandwidth greater than 110 GHz at 1550 nm. The design does not require etching of thin-film LN, and uses hybrid optical modes formed by bonding LN to planarized silicon photonic waveguide circuits. A high optical power handling capability in the MZM was achieved by carefully tapering the underlying Si waveguide to reduce the impact of optically-generated carriers, while retaining a high modulation efficiency. The MZM has a $V_πL$ product of 3.1 V$.$cm and an on-chip optical insertion loss of 1.8 dB.

physics.optics↗

arXiv2020

Meta R-CNN : Towards General Solver for Instance-level Few-shot Learning

Resembling the rapid learning capability of human, few-shot learning empowers vision systems to understand new concepts by training with few samples. Leading approaches derived from meta-learning on images with a single visual object. Obfuscated by a complex background and multiple objects in one image, they are hard to promote the research of few-shot object detection/segmentation. In this work, we present a flexible and general methodology to achieve these tasks. Our work extends Faster /Mask R-CNN by proposing meta-learning over RoI (Region-of-Interest) features instead of a full image feature. This simple spirit disentangles multi-object information merged with the background, without bells and whistles, enabling Faster /Mask R-CNN turn into a meta-learner to achieve the tasks. Specifically, we introduce a Predictor-head Remodeling Network (PRN) that shares its main backbone with Faster /Mask R-CNN. PRN receives images containing few-shot objects with their bounding boxes or masks to infer their class attentive vectors. The vectors take channel-wise soft-attention on RoI features, remodeling those R-CNN predictor heads to detect or segment the objects that are consistent with the classes these vectors represent. In our experiments, Meta R-CNN yields the state of the art in few-shot object detection and improves few-shot object segmentation by Mask R-CNN.

cs.CV↗

arXiv2019

Multivariate-Information Adversarial Ensemble for Scalable Joint Distribution Matching

A broad range of cross-$m$-domain generation researches boil down to matching a joint distribution by deep generative models (DGMs). Hitherto algorithms excel in pairwise domains while as $m$ increases, remain struggling to scale themselves to fit a joint distribution. In this paper, we propose a domain-scalable DGM, i.e., MMI-ALI for $m$-domain joint distribution matching. As an $m$-domain ensemble model of ALIs \cite{dumoulin2016adversarially}, MMI-ALI is adversarially trained with maximizing Multivariate Mutual Information (MMI) w.r.t. joint variables of each pair of domains and their shared feature. The negative MMIs are upper bounded by a series of feasible losses that provably lead to matching $m$-domain joint distributions. MMI-ALI linearly scales as $m$ increases and thus, strikes a right balance between efficacy and scalability. We evaluate MMI-ALI in diverse challenging $m$-domain scenarios and verify its superiority.

cs.LG↗

arXiv2017

Silicon Photonic Entangled Photon-Pair and Heralded Single Photon Generation with CAR $>$ 12,000 and $g^{(2)}(0)<$ 0.006

We report measurements of time-frequency entangled photon pairs and heralded single photons at 1550~nm wavelengths generated using a microring resonator pumped optically by a diode laser. Along with a high spectral brightness of pair generation, the conventional metrics used to describe performance, such as Coincidences-to-Accidentals Ratio (CAR), conditional self-correlation [$g^{(2)}(0)$], two-photon energy-time {F}ranson interferometric visibility etc. are shown to reach a high-performance regime not yet achieved by silicon photonics, and attained previously only by crystal, glass and fiber-based pair-generation devices.

quant-ph↗

arXiv2015

Recognize Foreign Low-Frequency Words with Similar Pairs

Low-frequency words place a major challenge for automatic speech recognition (ASR). The probabilities of these words, which are often important name entities, are generally under-estimated by the language model (LM) due to their limited occurrences in the training data. Recently, we proposed a word-pair approach to deal with the problem, which borrows information of frequent words to enhance the probabilities of low-frequency words. This paper presents an extension to the word-pair method by involving multiple `predicting words' to produce better estimation for low-frequency words. We also employ this approach to deal with out-of-language words in the task of multi-lingual speech recognition.

cs.CL↗