arXiv subjects

Xian Zhang

Publications and source records attributed to Xian Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2025

TrainVerify: Equivalence-Based Verification for Distributed LLM Training

Training large language models (LLMs) at scale requires parallel execution across thousands of devices, incurring enormous computational costs. Yet, these costly distributed trainings are rarely verified, leaving them prone to silent errors and potentially wasting millions of GPU hours. We introduce TrainVerify, a system for verifiable distributed training of LLMs. Given a deep learning model's logical specification as the ground truth, TrainVerify formally verifies that a distributed parallel execution plan is mathematically equivalent to it. Direct verification is notoriously difficult due to the sheer scale of LLMs which often involves billions of variables and highly intricate computation graphs. Therefore, TrainVerify introduces shape-reduction techniques and a stage-wise parallel verification algorithm that significantly reduces complexity while preserving formal correctness. TrainVerify scales to frontier LLMs, including the successful verification of the Llama3 (405B) and DeepSeek-V3 (671B) training plans.

cs.DC↗

arXiv2025

Reviving DSP for Advanced Theorem Proving in the Era of Reasoning Models

Recent advancements, such as DeepSeek-Prover-V2-671B and Kimina-Prover-Preview-72B, demonstrate a prevailing trend in leveraging reinforcement learning (RL)-based large-scale training for automated theorem proving. Surprisingly, we discover that even without any training, careful neuro-symbolic coordination of existing off-the-shelf reasoning models and tactic step provers can achieve comparable performance. This paper introduces \textbf{DSP+}, an improved version of the Draft, Sketch, and Prove framework, featuring a \emph{fine-grained and integrated} neuro-symbolic enhancement for each phase: (1) In the draft phase, we prompt reasoning models to generate concise natural-language subgoals to benefit the sketch phase, removing thinking tokens and references to human-written proofs; (2) In the sketch phase, subgoals are autoformalized with hypotheses to benefit the proving phase, and sketch lines containing syntactic errors are masked according to predefined rules; (3) In the proving phase, we tightly integrate symbolic search methods like Aesop with step provers to establish proofs for the sketch subgoals. Experimental results show that, without any additional model training or fine-tuning, DSP+ solves 80.7\%, 32.8\%, and 24 out of 644 problems from miniF2F, ProofNet, and PutnamBench, respectively, while requiring fewer budgets compared to state-of-the-arts. DSP+ proves \texttt{imo\_2019\_p1}, an IMO problem in miniF2F that is not solved by any prior work. Additionally, DSP+ generates proof patterns comprehensible by human experts, facilitating the identification of formalization errors; For example, eight wrongly formalized statements in miniF2F are discovered. Our results highlight the potential of classical reasoning patterns besides the RL-based training. All components will be open-sourced.

cs.AI↗

arXiv2025

The Future of the Correlated Electron Problem

A central problem in modern condensed matter physics is the understanding of materials with strong electron correlations. Despite extensive work, the essential physics of many of these systems is not understood and there is very little ability to make predictions in this class of materials. In this manuscript we share our personal views on the major open problems in the field of correlated electron systems. We discuss some possible routes to make progress in this rich and fascinating field. This manuscript is the result of the vigorous discussions and deliberations that took place at Johns Hopkins University during a three-day workshop January 27, 28, and 29, 2020 that brought together six senior scientists and 46 more junior scientists. Our hope, is that the topics we have presented will provide inspiration for others working in this field and motivation for the idea that significant progress can be made on very hard problems if we focus our collective energies.

cond-mat.str-el↗

arXiv2025

Jets from a stellar-mass black hole are as relativistic as those from supermassive black holes

Relativistic jets from supermassive black holes in active galactic nuclei are amongst the most powerful phenomena in the universe, acting to regulate the growth of massive galaxies. Similar jets from stellar-mass black holes offer a chance to study the same phenomena on accessible observation time scales. However, such comparative studies across black hole masses and time scales remain hampered by the long-standing perception that stellar-mass black hole jets are in a less relativistic regime. We used radio interferometry observations to monitor the Galactic black hole X-ray binary 4U 1543-47 and discovered two distinct, relativistic ejections launched during a single outburst. Our measurements reveal a likely Lorentz factor of $\sim$ 8 and a minimum of 4.6 at launch with 95% confidence, demonstrating that stellar-mass black holes in X-ray binaries can launch jets as relativistic as those seen in active galactic nuclei.

astro-ph.HE↗

arXiv2025

Quantum Reinforcement Learning-Based Two-Stage Unit Commitment Framework for Enhanced Power Systems Robustness

Unit commitment (UC) optimizes the start-up and shutdown schedules of generating units to meet load demand while minimizing costs. However, the increasing integration of renewable energy introduces uncertainties for real-time scheduling. Existing solutions face limitations both in modeling and algorithmic design. At the modeling level, they fail to incorporate widely adopted virtual power plants (VPPs) as flexibility resources, missing the opportunity to proactively mitigate potential real-time imbalances or ramping constraints through foresight-seeing decision-making. At the algorithmic level, existing probabilistic optimization, multi-stage approaches, and machine learning, face challenges in computational complexity and adaptability. To address these challenges, this study proposes a novel two-stage UC framework that incorporates foresight-seeing sequential decision-making in both day-ahead and real-time scheduling, leveraging VPPs as flexibility resources to proactively reserve capacity and ramping flexibility for upcoming renewable energy uncertainties over several hours. In particular, we develop quantum reinforcement learning (QRL) algorithms that integrate the foresight-seeing sequential decision-making and scalable computation advantages of deep reinforcement learning (DRL) with the parallel and high-efficiency search capabilities of quantum computing. Experimental results demonstrate that the proposed QRL-based approach outperforms in computational efficiency, real-time responsiveness, and solution quality.

eess.SY↗

arXiv2025

Simultaneous multi-wavelength observations of the repeating fast radio burst FRB 20190520B with Swift and FAST

Among several dozen known repeating Fast radio bursts (FRBs), those precisely localized offer the best opportunities to explore their multi-wavelength counterparts, which are key to uncovering their origins. Here we report our X-ray, ultraviolet (UV), and optical observations with the $Swift$ satellite of the repeating FRB 20190520B, in coordination with simultaneous radio observations with the Five-hundred-meter Aperture Spherical radio Telescope (FAST). Our aim was to detect potentially associated multi-wavelength bursts and identify any persistent multi-wavelength counterpart to the associated persistent radio source (PRS). While a total of 10 radio bursts were detected by FAST during the $Swift$ observations, we detected no X-ray, UV, or optical bursts in accompany with the radio bursts. We obtained the energy upper limits ($3σ$) on any multi-wavelength bursts as follows: $5.03 \times 10^{47}$ erg in the hard X-ray band (15--150 keV), $7.98 \times 10^{45}$ erg in the soft X-ray band (0.3--10 keV), and $4.51 \times 10^{44}$ erg in the U band (3465Å), respectively. The energy ratio between soft X-ray (0.3--10 keV) and radio emission of the bursts is constrained as $<6\times10^{7}$, and the ratio between optical (U band) and radio as $<1.19\times10^{6}$. We detect no multi-wavelength counterpart to the PRS. The 3$σ$ luminosity upper limits are 1.04$\times10^{47}$ (15--150 keV), 8.81$\times10^{42}$ (0.3--10 keV), 9.26$\times10^{42}$ (UVW1), and 2.54$\times10^{42}$ erg s$^{-1}$ (U), respectively. We show that the PRS is much more radio loud than representative pulsar wind nebulae, supernova remnants, extended jets of Galactic X-ray binaries, and ultraluminous X-ray sources, suggestive of boosted radio emission of the PRS.

astro-ph.HE↗

arXiv2025

Proving Olympiad Inequalities by Synergizing LLMs and Symbolic Reasoning

Large language models (LLMs) can prove mathematical theorems formally by generating proof steps (\textit{a.k.a.} tactics) within a proof system. However, the space of possible tactics is vast and complex, while the available training data for formal proofs is limited, posing a significant challenge to LLM-based tactic generation. To address this, we introduce a neuro-symbolic tactic generator that synergizes the mathematical intuition learned by LLMs with domain-specific insights encoded by symbolic methods. The key aspect of this integration is identifying which parts of mathematical reasoning are best suited to LLMs and which to symbolic methods. While the high-level idea of neuro-symbolic integration is broadly applicable to various mathematical problems, in this paper, we focus specifically on Olympiad inequalities (Figure~1). We analyze how humans solve these problems and distill the techniques into two types of tactics: (1) scaling, handled by symbolic methods, and (2) rewriting, handled by LLMs. In addition, we combine symbolic tools with LLMs to prune and rank the proof goals for efficient proof search. We evaluate our framework on 161 challenging inequalities from multiple mathematics competitions, achieving state-of-the-art performance and significantly outperforming existing LLM and symbolic approaches without requiring additional training data.

cs.AI↗

arXiv2025

A flaring radio counterpart to a fast radio burst reveals a newborn magnetized engine

Fast Radio Bursts (FRBs) are energetic millisecond radio bursts at cosmological distances, whose underlying engine is not identified. Among a sub-population that emit repeated radio bursts, a handful were associated with a persistent radio source (PRS) whose origin is unknown. Here we report the discovery of a compact flaring radio source (FRS) associated with a newly-active repeating FRB within one month after the first radio burst was detected. Its temporal and spectral characteristics differ from those of the PRSs but are similar to those of engine-powered supernovae and low-luminosity active galactic nuclei. We detected a spectral peak around $1.6\pm0.2$ GHz that is consistent with synchrotron self-absorption. Assuming equipartition, the magnetic field strength in the FRS is larger than the line-of-sight component constrained from the FRB Faraday rotation, suggesting a highly magnetized engine. The radius of the FRS is constrained to be $\sim0.03$ pc and the minimum total energy is $\sim~6.2\times{10}^{47}~{\rm ergs~}$. This FRS reveals the birth of a highly magnetized FRB engine, and hints that PRSs associated with other active FRBs may be in the later stage of evolution.

astro-ph.HE↗

arXiv2025

Multi-task Visual Grounding with Coarse-to-Fine Consistency Constraints

Multi-task visual grounding involves the simultaneous execution of localization and segmentation in images based on textual expressions. The majority of advanced methods predominantly focus on transformer-based multimodal fusion, aiming to extract robust multimodal representations. However, ambiguity between referring expression comprehension (REC) and referring image segmentation (RIS) is error-prone, leading to inconsistencies between multi-task predictions. Besides, insufficient multimodal understanding directly contributes to biased target perception. To overcome these challenges, we propose a Coarse-to-fine Consistency Constraints Visual Grounding architecture ($\text{C}^3\text{VG}$), which integrates implicit and explicit modeling approaches within a two-stage framework. Initially, query and pixel decoders are employed to generate preliminary detection and segmentation outputs, a process referred to as the Rough Semantic Perception (RSP) stage. These coarse predictions are subsequently refined through the proposed Mask-guided Interaction Module (MIM) and a novel explicit bidirectional consistency constraint loss to ensure consistent representations across tasks, which we term the Refined Consistency Interaction (RCI) stage. Furthermore, to address the challenge of insufficient multimodal understanding, we leverage pre-trained models based on visual-linguistic fusion representations. Empirical evaluations on the RefCOCO, RefCOCO+, and RefCOCOg datasets demonstrate the efficacy and soundness of $\text{C}^3\text{VG}$, which significantly outperforms state-of-the-art REC and RIS methods by a substantial margin. Code and model will be available at \url{https://github.com/Dmmm1997/C3VG}.

cs.CV↗

arXiv2024

Reduction in Thermal Conductivity of Monolayer MoS2 by Large Mechanical Strains for Efficient Thermal Management

Two dimensional (2D) materials such as graphene and transition metal dichalcogenides (TMDC) have received extensive research interests and investigations in the past decade. In this research, we report the first experimental measurement of the in plane thermal conductivity of MoS2 monolayer under a large mechanical strain using optothermal Raman technique. This measurement technique is direct without additional processing to the material, and MoS2's absorption coefficient is discovered during the measurement process to further increase this technique's precision. Tunable uniaxial tensile strains are applied on the MoS2 monolayer by stretching a flexible substrate it sits on. Experimental results demonstrate that, the thermal conductivity is substantially suppressed by tensile strains: under the tensile strain of 6.3%, the thermal conductivity of the MoS2 monolayer drops approximately by 62%. A serious of thermal transport properties at a group of mechanical strains are also reported, presenting a strain dependent trend. It is the first and original study of 2D materials' thermal transport properties under a large mechanical strain, and provides important information that the thermal transport of MoS2 will significantly decrease at a large mechanical strain. This finding provides the key information for flexible and wearable electronics thermal management and designs.

cond-mat.mtrl-sci↗

arXiv2024

Layer Dependent Thermal Transport Properties of One- to Three-Layer Magnetic Fe:MoS2

Two-Dimensional (2D) transition metal dichalcogenides (TMDs) have been the subject of extensive attention thanks to their unique properties and atomically thin structure. Because of its unprecedented room-temperature magnetic properties, iron-doped MoS2 (Fe:MoS2) is considered the next-generation quantum and magnetic material. It is essential to understand Fe:MoS2's thermal behavior since temperature and thermal load/activation are crucial for their magnetic properties and the current nano and quantum devices have been severely limited by thermal management. In this work, Fe:MoS2 is synthesized by doping Fe atoms into MoS2 using the chemical vapor deposition (CVD) synthesis and a refined version of opto-thermal Raman technique is used to study the thermal transport properties of Fe:MoS2 in the forms of single (1L), bilayer (2L), and tri-layer (3L). In the Opto-thermal Raman technique, a laser is focused on the center of a thin film and used to measure the peak position of a Raman-active mode. The lateral thermal conductivity of 1-3L of Fe:MoS2 and the interfacial thermal conductance between Fe:MoS2 and the substrate were obtained by analyzing the temperature-dependent and power-dependent Raman measurement, laser power absorption coefficient, and laser spot sizes. We also characterized Fe:MoS2's thermal transport at high temperature, and calculated Fe:MoS2's thermal transport by density theory function. These findings will shed light on the thermal management and thermoelectric designs for Fe:MoS2 based nano and quantum electronic devices.

cond-mat.mtrl-sci↗

arXiv2024

Autoformalize Mathematical Statements by Symbolic Equivalence and Semantic Consistency

Autoformalization, the task of automatically translating natural language descriptions into a formal language, poses a significant challenge across various domains, especially in mathematics. Recent advancements in large language models (LLMs) have unveiled their promising capabilities to formalize even competition-level math problems. However, we observe a considerable discrepancy between pass@1 and pass@k accuracies in LLM-generated formalizations. To address this gap, we introduce a novel framework that scores and selects the best result from k autoformalization candidates based on two complementary self-consistency methods: symbolic equivalence and semantic consistency. Elaborately, symbolic equivalence identifies the logical homogeneity among autoformalization candidates using automated theorem provers, and semantic consistency evaluates the preservation of the original meaning by informalizing the candidates and computing the similarity between the embeddings of the original and informalized texts. Our extensive experiments on the MATH and miniF2F datasets demonstrate that our approach significantly enhances autoformalization accuracy, achieving up to 0.22-1.35x relative improvements across various LLMs and baseline methods.

cs.CL↗

arXiv2024

Neuro-Symbolic Data Generation for Math Reasoning

A critical question about Large Language Models (LLMs) is whether their apparent deficiency in mathematical reasoning is inherent, or merely a result of insufficient exposure to high-quality mathematical data. To explore this, we developed an automated method for generating high-quality, supervised mathematical datasets. The method carefully mutates existing math problems, ensuring both diversity and validity of the newly generated problems. This is achieved by a neuro-symbolic data generation framework combining the intuitive informalization strengths of LLMs, and the precise symbolic reasoning of math solvers along with projected Markov chain Monte Carlo sampling in the highly-irregular symbolic space. Empirical experiments demonstrate the high quality of data generated by the proposed method, and that the LLMs, specifically LLaMA-2 and Mistral, when realigned with the generated data, surpass their state-of-the-art counterparts.

cs.AI↗

arXiv2024

Flaring gamma-ray emission coincident with a hyperactive fast radio burst source

Fast radio bursts (FRBs) are bright milliseconds-duration radio bursts from cosmological distances. Despite intense observational and theoretical studies, their physical origin is still mysterious. One major obstacle is the lack of identification of multi-wavelength counterparts for FRBs at cosmological distances. So far, all the searches other than in the radio wavelength, including those in the gamma-ray energies, have only left upper limits. Here we report a gigaelectronvolt (GeV) gamma-ray flare lasting 15.6 seconds as well as additional evidence of variable gamma-ray emission in temporal and spatial association with the hyper-active, newly discovered repeating FRB 20240114A, which has been localized to a dwarf galaxy at a redshift of 0.13. The energetic, short GeV gamma-ray flare reached a prompt isotropic luminosity of the order of ${10}^{48}~{\rm ergs~{s}^{-1}}$. The additional less-significant gamma-ray flares, if true, also have similar luminosities; such flares could contribute to a 5-day average luminosity of the order of ${10}^{45}~{\rm ergs~{s}^{-1}}$. These high-luminosity flares challenge the traditional FRB engine scenario involving a seconds-period magnetar. Rather, it suggests a powerful, long-lived, but newborn energy source at the location of this active repeater, either directly powering the bursts or indirectly triggering bursts in the vicinity of the FRB engine.

astro-ph.HE↗

arXiv2024

Conceiving Naturally After IVF: the effect of assisted reproduction on obstetric interventions and child health at birth

A growing share of the world's population is being born via assisted reproductive technology (ART), including in-vitro fertilisation (IVF). However, two concerns persist. First, ART pregnancies correlate with predictors of poor outcomes at birth--and it is unclear whether this relationship is causal. Second, the emotional and financial costs associated with ART-use might exacerbate defensive medical behaviour, where physicians intervene more than necessary to reduce the risk of adverse medical outcomes and litigation. We address the challenge of identifying the pure effect of ART-use on both maternal and infant outcomes at birth by leveraging exogenous variation in the success of ART cycles. We compare the obstetric outcomes for ART-conceived births with those of spontaneously-conceived births after a failed ART treatment. Moreover, we flexibly adjust for key confounders using double machine learning. We do this using clinical registry ART data and administrative maternal and infant data from New South Wales (NSW) between 2009-2017. We find that ART slightly decreases the risk of obstetric interventions, lowering the risk of a caesarean section and increasing the rate of spontaneous labour (+3.5 p.p.). Moreover, we find that ART has a statistically and clinically insignificant effect on infant health outcomes. Keywords: Fertility, Assisted reproduction, IVF, Caesarean Section, Obstetric, Infertility. JEL classification: I10, I12, I19.

econ.GN↗

arXiv2024

A Survey on Deep Learning for Theorem Proving

Theorem proving is a fundamental aspect of mathematics, spanning from informal reasoning in natural language to rigorous derivations in formal systems. In recent years, the advancement of deep learning, especially the emergence of large language models, has sparked a notable surge of research exploring these techniques to enhance the process of theorem proving. This paper presents a comprehensive survey of deep learning for theorem proving by offering (i) a thorough review of existing approaches across various tasks such as autoformalization, premise selection, proofstep generation, and proof search; (ii) an extensive summary of curated datasets and strategies for synthetic data generation; (iii) a detailed analysis of evaluation metrics and the performance of state-of-the-art methods; and (iv) a critical discussion on the persistent challenges and the promising avenues for future exploration. Our survey aims to serve as a foundational reference for deep learning approaches in theorem proving, inspiring and catalyzing further research endeavors in this rapidly growing field. A curated list of papers is available at https://github.com/zhaoyu-li/DL4TP.

cs.AI↗

arXiv2024

Non-Linear behavior of Raman Linewidth of 1-3 Layer WSe2

In recent years, Raman spectroscopy of two-dimensional (2D) materials has been conducted. However, only a small portion of those studies note the full width half maximum (FWHM) of these 2D materials. FWHM is an indicator of phonon dispersions in 2D materials. In addition, it is a critical parameter that can denote the quality of the given 2D material, which is directly related to its engineering performance for manufactured devices. In this study, we analyze the effect that temperature has on the FWHM of the E2g1 mode for WSe2. We also analyze this temperature dependent trend for WSe2 with different thicknesses: monolayer (1L), bilayer (2L), and triple-layer (3L). The results show a different trend associated with each thickness on temperature dependence analysis for WSe2. The 1L WSe2 shows a parabolic trend with increasing temperature. Meanwhile, the 2L shows a steady linear trend with increasing temperature. On the other hand, FWHM for 3L WSe2 can be fitted by a simplified Klemens model, which explains its non-linear behavior.

physics.optics↗

arXiv2024

Nonvolatile spin field effect transistor based on VSi2N4/Sc2CO2 multiferroic heterostructure

In this study, we present first-principles calculations that introduce a novel nonvolatile spin field-effect transistor (Spin-FET) utilizing a van der Waals multiferroic heterostructure, specifically VSi2N4/Sc2CO2. We demonstrate that inverting the ferroelectric polarization in a Sc2CO2 monolayer can effectively modulate the electronic states of a VSi2N4 monolayer, enabling a transition from half-metal to half-semiconductor. This transition significantly alters the electronic transport properties. Furthermore, we construct a Spin-FET device based on this multiferroic heterostructure and observe that the VSi2N4/Sc2CO2-based Spin-FET exhibits exceptional all-electric-controlled performance. Notably, the inversion of the Sc2CO2 ferroelectric polarization yields a substantial on-off current ratio, approximately 650\%, under a minimal bias voltage of 0.02 V. Additionally, we identify a unique spatially-separated spin-polarized transport phenomenon, wherein pure spin-up electrons transport exclusively through VSi2N4, and spin-down electrons through Sc$_2$CO$_2$. Our findings suggest a promising pathway for developing low-energy-dissipation and nonvolatile FET devices.

cond-mat.mes-hall↗