arXiv subjects

Biao Tang

Publications and source records attributed to Biao Tang.

18 recordsLinked to original sources

arXiv2026

DualEvolve: Dual-Level Self-Evolution of Mechanism and Orchestration for Recommendation

Recommender systems are widely used in e-commerce and content platforms, where improvement requires repeated model design, training optimization, and experimental validation. Large language model (LLM) agents increasingly automate this development cycle. Sustained autonomous improvement requires agents to improve both recommendation models and their development workflows. However, task-performance-driven approaches select updates mainly by candidates' final scores, without a separate optimization signal for experimentation quality. We therefore present DualEvolve, a dual-level framework that co-evolves recommendation mechanisms and agent orchestration toward recursive self-improvement. Inner evolution revises model and training code using task feedback to improve recommendation performance. Outer evolution scores the inner-loop experimental process, updates harness and skill rules using these scores and execution traces, and adjusts orchestration choices based on next-round scores to improve the quality of subsequent experiments. Experiments on MovieLens-1M and three Amazon Reviews 2023 subsets cover four recommendation methods and 16 settings. DualEvolve achieves higher test means than Base, OpenEvolve, and DeepEvolve in all 64 setting-metric comparisons, with macro-averaged relative gains over DeepEvolve of 7.6% (HR@5) and 9.0% (NDCG@5).

cs.IR↗

arXiv2026

SPADE: Spatiotemporal Preference Alignment and Diversity Enhancement for Local Search

Proactive query recommendation aims to suggest relevant search queries to users on the homepage of local service apps. Unlike conventional recommendation over a fixed item set, this requires open-set generation from a vocabulary of millions, where outputs must simultaneously align with user preference, match fine-grained spatiotemporal context, cover long-tail intents, and provide diverse alternatives within a few candidate slots. Existing approaches address these axes in isolation: standard tokenization suppresses long-tail queries, coarse location encoding misses fine-grained geographic demand, generic preference optimization fails to align with production impression-click distributions, and standard beam search collapses to near-duplicate outputs. We present SPADE (Spatiotemporal Preference Alignment and Diversity Enhancement), an end-to-end generative framework that jointly optimizes the generation space, contextual representation, preference distribution, and output slate quality through four components: Objective-Aligned BPE (OA-BPE) aligns subword merging with business value signals to boost long-tail query generation probability; Spatio-Temporal Adaptive Representation (STAR) replaces coarse location tokens with continuous multi-scale encodings and soft gating for precise spatiotemporal grounding; MARS-DPO (Multi-dimensional Adaptive Reward Sampling DPO) combines cross-domain pre-training with preference alignment on production impression-click data via composite rewards and adaptive hard negative sampling; and Sigmoid Diversity Beam Search (SDBS) improves intent coverage without sacrificing recall. Offline experiments on large-scale production data show that SPADE outperforms the strongest generative baseline by +5.6% HR@5, -40.3% GeoC@5, and +1.9% Exec@5. A one-week online A/B test confirms production impact: +1.88% Module CTR, +0.177% Search Volume, and +0.105% Search DAU.

cs.IR↗

arXiv2026

Learning and Forecasting Multiscale Spatiotemporal Epidemic Dynamics with Lagrangian-Eulerian Graph Neural ODEs

Accurate medium- to long-range epidemic forecasting requires extrapolating high-dimensional, heterogeneous and nonstationary dynamics from partial observations. Yet hybrid models often leave implicit which observations inform which epidemic processes and how learned graph interactions modify system state. We introduce LE-EpiGNODE, which replaces generic message passing with process-specific, mechanism-guided neural operators: age-contact information drives Lagrangian infection generation, whereas mobility drives conservative Eulerian population transport within a structure-preserving universal differential equation. We evaluated LE-EpiGNODE across 14 controlled synthetic mechanisms spanning time-modulated, state-dependent, nonlinear and coupled dynamics, three complementary COVID-19 settings, 21 comparator models and five targeted ablations. Forecasting gains tended to become more pronounced at longer horizons; at the longest horizon, LE-EpiGNODE achieved the lowest WMAPE in all three real-world datasets, reducing error by 17.7-40.0% relative to the best baseline. Controlled ablations linked these gains to the Lagrangian-Eulerian decomposition and process-aligned priors, while synthetic ground truth showed that learned operators and induced state fluxes track known generating dynamics. These results support mechanism-organized graph neural dynamics as a route to reliable long-horizon extrapolation with directly testable, process-level mechanistic interpretation beyond trajectory accuracy.

cs.LG↗

arXiv2026

ReUnit: Multi-Granularity Visual Unitization for Long Video Understanding

Long-video understanding is constrained by the limited visual input capacity of video multimodal large language models (Video-MLLMs). Existing methods mainly optimize which content to retain, while the presentation of retained content often remains fixed. As a result, the same balance between spatial detail and content coverage is imposed across the entire visual input. We propose ReUnit, a training-free and query-aware framework that jointly determines which content to retain and how it should be presented. Guided by frame-level query relevance, ReUnit constructs and allocates visual units at multiple presentation granularities. More relevant content receives finer presentation, while broader context is represented more compactly. It realizes these granularities using visual units that carry one, four, or nine source frames and are each rendered as a standard image. Across four benchmarks and visual input budgets from 4 to 32 units, ReUnit achieves the highest average score at every tested budget. It improves over Uniform Sampling by 8.3--10.1 points on average, and the gains transfer across three additional Video-MLLM families. Project resources are available at https://github.com/charon525/ReUnit.

cs.CV↗

arXiv2026

MMUSV-Sim: A Perception-Oriented Simulation and Data-Generation Platform for Multi-USV Cooperative Perception

Cooperative perception among multiple unmanned surface vehicles (USVs) combines complementary observations to extend maritime target sensing beyond the view range and field of a single platform. Developing such systems at scale calls for a unified workflow for configurable multi-USV scenarios, multimodal acquisition, and shared annotations. We present MMUSV-Sim, a perception-oriented maritime simulation and data-generation platform built on Unreal Engine 5 and Project AirSim. It provides island, open-sea, and port environments; configurable weather, time of day, and wave conditions; a diverse vessel asset library; and spline-based multi-vessel motion. MMUSV-Sim acquires RGB, depth, semantic, LiDAR, and radar observations across multiple USVs and captures a common world state for per-agent annotation export. Experiments verify that the configured wave settings produce the intended changes in vessel heave, roll, and pitch, and evaluate the geometric consistency between projected annotations and semantic renderings. In LiDAR-based cooperative BEV vessel detection experiments on the generated multi-USV dataset, Early Fusion achieves an AP@0.5 of 72.74, compared with 45.54 using a single USV.

cs.RO↗

arXiv2026

Amplitude Modulation Noise Suppression of Dynamic Atom Gravimeters

Dynamic atom gravimeters enable absolute gravity measurements on moving platforms. However, their performance is severely degraded due to the complex dynamic environment. This paper finds that the amplitude modulation noise (AMN) is a key factor contributing to the degradation of gravity measurement performance. We find that the AMN is induced by the cold atomic cloud trajectory and velocity variation. We build a model to illustrate the principles and magnitude of AMN arising from various experiment processes. Then we propose a method to fit the normalized AMN respect to the kinematic parameters of the cold atomic cloud, and successfully suppress this noise from 0.11 to 0.038 using the fitting result. With this method, we improve the fringe phase resolution from 0.244 rad to 0.092 rad, and reduce the dynamic gravity measurement noise from 2.69 mGal to 1.68 mGal. This study finds and suppresses a key noise source for the dynamic atom gravimeters, which is important for further improving its precision. The proposed method can be also applied for precision enhancement for other dynamic atom interferometer-based sensors, such as the atom gradiometers and gyroscopes.

physics.atom-ph↗

arXiv2026

In-orbit Test of the Weak Equivalence Principle with Atom Interferometry

The Weak Equivalence Principle (WEP) is a central pillar of general relativity. Its precise test with quantum systems in space offers a unique window onto new physics. Here we report the first in-orbit quantum test of the WEP. A dual-species (85Rb/87Rb) atom interferometer is realized aboard the China Space Station. Methods of platform motion suppression, fluorescence detection switching, and two-photon detuning switching are developed to eliminate phase noise and improve measurement accuracy. A test uncertainty of 2.8*10-8 is obtained from 280 days of WEP test data, and a test result of (-3.1+/-4.6)*10-7 is achieved after error estimation. This improves prior atom-interferometric WEP tests in microgravity by three orders of magnitude. This work paves the way for space-borne quantum inertial sensors and their application to future fundamental physics in space.

physics.atom-ph↗

arXiv2025

Robust Atom Interferometry with Super-Gaussian Pulses against Thermal Velocity Spread

Laser frequency fluctuation and atomic thermal motion can lead to errors in pulse duration and detuning in cold atom interferometry, thereby reducing measurement stability and fringe contrast. To address this issue, we investigate the use of super-Gaussian pulses, which are characterized by smooth temporal profiles and centralized energy distribution, in the beam-splitting and reflection stages of an atom interferometer. Through numerical simulations, we compare the performance of rectangular, Gaussian, and 2nd- to 10th-order super-Gaussian pulses subject to deviations in pulse duration and detuning. Our results show that both Gaussian and super-Gaussian pulses offer a significant advantage over traditional rectangular pulses, particularly under thermal conditions where velocity spread is prominent. We find that 4th-order pulses achieving up to a 90\% improvement in contrast over rectangular pulses under realistic conditions, and while their peak performance at very low temperatures is comparable to that of Gaussian pulses, they demonstrate enhanced robustness against combined detuning and pulse-length errors. These findings demonstrate that super-Gaussian pulse shaping is an effective method for enhancing the robustness of atom interferometers against errors induced by thermal motion.

physics.atom-ph↗

arXiv2025

Analysis of the Dick Effect for AI-based Dynamic Gravimeter

Atom interferometer (AI)-based dynamic gravimeter enable high-precision absolute gravity measurements, crucial for applications in geophysics, navigation, resource exploration, and metrology. Understanding their underlying mechanisms and minimizing measurement noise are essential for enhancing performance. This work investigates the gravity measurement noise in AI-based systems induced by the dead time of the classical accelerometer. Using actual dynamic gravity measurement data, we demonstrate that a dead time of 0.12 s introduces significant gravity measurement noise of 8 mGal. To elucidate the mechanism of this noise, we derive a formula for this noise in frequency domain, identifying high-frequency aliasing as its source. Analysis of the derived expressions indicates that reducing the dead time duration and suppressing the high-frequency noise of the acceleration are effective strategies for mitigating this noise. This work provides significant insights for noise analysis and future scheme design of AI-based dynamic gravimeters.

physics.atom-ph↗

arXiv2025

Compact and robust design of the optical system for cold atom interferometer in space

The optical system is a complex and precise subsystem for the atom interferometer (AI), especially for those used in field or space applications. Here, we introduce the design of the optical system of the China Space Station atom interferometer (CSSAI). The scheme is optimized to reduce the complexity while maintaining the capability to achieve the dual-species AI. It features a fused silica optical bench with bonding technology, ensuring compactness and smaller thermal deformation. Spatial structures are designed to isolate the vibration and transfer the heat. After assembling, the optical system has a size of 250 mm * 240 mm * 104 mm and weighs 5.2 kg. After installing in the CSSAI, it passed the thermal and mechanical tests and then launched to the China Space Station (CSS). The output laser power changes are less than 15% from ground to space, and its long-term fluctuations are less than 2.5% for months in space. Cold atom preparation and interference are also realized in space. This optical system is extremely integrated and robust, which provides a foundation for the design of future cold atom payloads in space.

physics.atom-ph↗

arXiv2025

Detection of subsurface structures with a vehicle-based atom gravity gradiometer

High-precision mobile gravity gradiometers are very useful in geodesy and geophysics. Atom gravity gradiometers (AGGs) could be among the most accurate mobile gravity gradiometers but are currently constrained by the trade-off between portability and sensitivity. Here, we present a high-sensitivity mobile AGG featuring an ultra-compact sensor head with a volume of only 94 L. In the laboratory, it achieves a sensitivity of 77 E/$\sqrt{Hz}$ (1 E=1$\times10^{-9}$/s$^2$) and a long-term stability of better than 0.5 E. We integrated the instrument in a minivan, enabling efficient mobile field surveys with excellent maneuverability in confined spaces. Using this vehicular system, we surveyed the gravitational field over a set of subsurface structures within a small wooded area, successfully resolving their structural signatures with a signal-to-noise ratio of 57 and quantifying the water depth in a reservoir with an accuracy of $\pm$0.23 m. Compared with previous observations using a CG-5 gravimeter, the superior spatial resolution inherent in gradiometry is clearly demonstrated. This work paves the way for bring AGGs to practical field applications.

physics.ins-det↗

arXiv2025

Personalized Image Generation with Large Multimodal Models

Personalized content filtering, such as recommender systems, has become a critical infrastructure to alleviate information overload. However, these systems merely filter existing content and are constrained by its limited diversity, making it difficult to meet users' varied content needs. To address this limitation, personalized content generation has emerged as a promising direction with broad applications. Nevertheless, most existing research focuses on personalized text generation, with relatively little attention given to personalized image generation. The limited work in personalized image generation faces challenges in accurately capturing users' visual preferences and needs from noisy user-interacted images and complex multimodal instructions. Worse still, there is a lack of supervised data for training personalized image generation models. To overcome the challenges, we propose a Personalized Image Generation Framework named Pigeon, which adopts exceptional large multimodal models with three dedicated modules to capture users' visual preferences and needs from noisy user history and multimodal instructions. To alleviate the data scarcity, we introduce a two-stage preference alignment scheme, comprising masked preference reconstruction and pairwise preference alignment, to align Pigeon with the personalized image generation task. We apply Pigeon to personalized sticker and movie poster generation, where extensive quantitative results and human evaluation highlight its superiority over various generative baselines.

cs.IR↗

arXiv2023

High precision atom interferometer-based dynamic gravimeter measurement by eliminating the cross-coupling effect

A dynamic gravimeter with an atomic interferometer (AI) can perform absolute gravity measurements with high precision. AI-based dynamic gravity measurement is a type of joint measurement that uses AI sensors and a classical accelerometer. The coupling of the two sensors may degrade the measurement precision. In this study, we analyzed the cross-coupling effect and introduced a recovery vector to suppress this effect. We improved the phase noise of the interference fringe by a factor of 1.9 by performing marine gravity measurements using an AI-based gravimeter and optimizing the recovery vector. Marine gravity measurements were performed, and high gravity measurement precision was achieved. The external and inner coincidence accuracies of the gravity measurement are 0.42 mGal and 0.46 mGal, which were improved by factors of 4.18 and 4.21 by optimizing the cross-coupling effect.

physics.app-ph↗

arXiv2022

Development of a compact high-resolution absolute gravity gradiometer based on atom interferometers

We present a compact high-resolution gravity gradiometer based on dual Rb-85 atom interferometers using stimulated Raman transitions. A baseline L=44.5 cm and an interrogation time T=130 ms are realized in a sensor head with volume of less than 100 liters. Experimental parameters are optimized to improve the short-term sensitivity while a rejection algorithm relying on inversion of the Raman wave vector is implemented to improve the long-term stability. After an averaging time of 17000 s, a phase resolution of 104 μrad is achieved, which corresponds to a gravity gradient resolution of 0.86 E. As far as we know, this is the sub-E atom gravity gradiometer with the highest level of compactness to date. After the evaluation and correction of system errors induced by light shift, residual Zeeman shift, Coriolis effect and self-attraction effect, the instrument serves as an absolute gravity gradiometer and with it the local gravity gradient is measured to be 3114 (53) E.

physics.atom-ph↗

arXiv2019

ZAIGA: Zhaoshan Long-baseline Atom Interferometer Gravitation Antenna

The Zhaoshan long-baseline Atom Interferometer Gravitation Antenna (ZAIGA) is a new type of underground laser-linked interferometer facility, and is currently under construction. It is in the 200-meter-on-average underground of a mountain named Zhaoshan which is about 80 km southeast to Wuhan. ZAIGA will be equipped with long-baseline atom interferometers, high-precision atom clocks, and large-scale gyros. ZAIGA facility will take an equilateral triangle configuration with two 1-km-apart atom interferometers in each arm, a 300-meter vertical tunnel with atom fountain and atom clocks mounted, and a tracking-and-ranging 1-km-arm-length prototype with lattice optical clocks linked by locked lasers. The ZAIGA facility will be used for experimental research on gravitation and related problems including gravitational wave detection, high-precision test of the equivalence principle of micro-particles, clock based gravitational red-shift measurement, rotation measurement and gravito-magnetic effect.

physics.atom-ph↗

arXiv2015

Influence of separating distance between atomic sensors for gravitational wave detection

We consider a recent scheme of gravitational wave detection using atomic interferometers as inertial sensors, and reinvestigate its configuration using the concept of sensitivity functions. We show that such configuration can suppress noise without influencing the gravitational wave signal. But the suppression is insufficient for the direct observation of gravitational wave signals, so we analyse the behaviour of the different noises influencing the detection scheme. As a novel method, we study the relations between the measurement sensitivity and the distance between two interferometers, and find that the results derived from vibration noise and laser frequency noise are in stark contrast to that derived from the shot noise, which is significant for the configuration design of gravitational wave detectors using atomic interferometers.

gr-qc↗

arXiv2015

Test of Equivalence Principle at $10^{-8}$ Level by a Dual-species Double-diffraction Raman Atom Interferometer

We report an improved test of the weak equivalence principle by using a simultaneous $^{85}$Rb-$^{87}$Rb dual-species atom interferometer. We propose and implement a four-wave double-diffraction Raman transition scheme for the interferometer, and demonstrate its ability in suppressing common-mode phase noise of Raman lasers after their frequencies and intensity ratios are optimized. The statistical uncertainty of the experimental data for Eötvös parameter $η$ is $0.8\times10^{-8}$ at 3200 s. With various systematic errors corrected the final value is $η=(2.8\pm3.0)\times10^{-8}$. The major uncertainty is attributed to the Coriolis effect.

physics.atom-ph↗

arXiv2013

Configurations of a new atomic interferometer for gravitational wave detection

Recently, the configuration using atomic interferometers (AIs) had been suggested for the detection of gravitational waves. A new AI with some additional laser pulses for implementing large momentum transfer was also put forward, in order to improve the influence of shot noise and laser frequency noise. In the paper, we use the sensitivity function to analyze all possible configurations of the new AI and to distinguish how many momenta are transferred in a specific configuration. With the analysis for the new configuration, we explore the detection scheme of gravitational wave further, in particular, for the amelioration of the laser frequency noise. We find that the amelioration is definite in such scheme, but novelly, in some cases the frequency noise can be canceled completely by using a proper data processing method.

gr-qc↗