arXiv subjects

Rui Shi

Publications and source records attributed to Rui Shi.

At least 55 records · Page 3Linked to original sources

arXiv2025

OmniSQL: Synthesizing High-quality Text-to-SQL Data at Scale

Text-to-SQL, the task of translating natural language questions into SQL queries, plays a crucial role in enabling non-experts to interact with databases. While recent advancements in large language models (LLMs) have significantly enhanced text-to-SQL performance, existing approaches face notable limitations in real-world text-to-SQL applications. Prompting-based methods often depend on closed-source LLMs, which are expensive, raise privacy concerns, and lack customization. Fine-tuning-based methods, on the other hand, suffer from poor generalizability due to the limited coverage of publicly available training data. To overcome these challenges, we propose a novel and scalable text-to-SQL data synthesis framework for automatically synthesizing large-scale, high-quality, and diverse datasets without extensive human intervention. Using this framework, we introduce SynSQL-2.5M, the first million-scale text-to-SQL dataset, containing 2.5 million samples spanning over 16,000 synthetic databases. Each sample includes a database, SQL query, natural language question, and chain-of-thought (CoT) solution. Leveraging SynSQL-2.5M, we develop OmniSQL, a powerful open-source text-to-SQL model available in three sizes: 7B, 14B, and 32B. Extensive evaluations across nine datasets demonstrate that OmniSQL achieves state-of-the-art performance, matching or surpassing leading closed-source and open-source LLMs, including GPT-4o and DeepSeek-V3, despite its smaller size. We release all code, datasets, and models to support further research.

cs.CL↗

arXiv2025

Leanabell-Prover-V2: Verifier-integrated Reasoning for Formal Theorem Proving via Reinforcement Learning

We introduce our Leanabell-Prover-V2, a 7B large language models (LLMs) that can produce formal theorem proofs in Lean 4, with verifier-integrated Long Chain-of-Thoughts (CoT). Following our previous work Leanabell-Prover-V1, we continual to choose to posttrain existing strong prover models for further performance improvement. In our V2 version, we mainly upgrade the Reinforcement Learning (RL) with feedback provided by the Lean 4 verifier. Crucially, verifier feedback, such as indicating success or detailing specific errors, allows the LLM to become ``self-aware'' of the correctness of its own reasoning process and learn to reflexively correct errors. Leanabell-Prover-V2 directly optimizes LLM reasoning trajectories with multi-turn verifier interactions, together with feedback token masking for stable RL training and a simple reward strategy. Experiments show that Leanabell-Prover-V2 improves performance by 3.2% (pass@128) with Kimina-Prover-Preview-Distill-7B and 2.0% (pass@128) with DeepSeek-Prover-V2-7B on the MiniF2F test set. The source codes, curated data and models are available at: https://github.com/Leanabell-LM/Leanabell-Prover-V2.

cs.AI↗

arXiv2025

Toward Sustainable Rare Earth Element Production: Key Challenges in Techno-Economic, Life Cycle, and Social Impact Assessment

Rare earth elements (REEs) are 17 critical minerals used in many clean energy technologies like wind turbines and electric vehicles. Conventionally, we produce REEs from mining in few, geopolitically restricted regions. Developing systems that utilize new technologies and unconventional feedstocks provides an opportunity to meet increasing demand while improving sustainability. Techno-economic analysis (TEA), life cycle assessment (LCA), and social LCA (sLCA) are commonly used tools to assess the sustainability performance of these systems. However, analyses of REE systems encounter challenges including system scope, data availability, technology scale-up, and uncertainty. In the reviewed literature, systems served multiple functions beyond producing REEs, including circularizing production and waste remediation, leading to discrepancies in scope. Further, the instability of REE prices led to high uncertainty due to different revenue, costs, functional unit, and impact allocation. Therefore, these analyses leave decision makers with an incomplete understanding of the current landscape of REE production inhibiting intelligent and efficient identification of future direction. In this narrative review, we conducted a comprehensive overview of the literature, synthesized studies from each pillar of sustainability (economic, environmental, and social), highlighted the challenges and limitations in each field, and recommended direction for future work developing sustainable REE production systems.

physics.soc-ph↗

arXiv2025

A Measurement of the Largest-Scale CMB E-mode Polarization with CLASS

We present measurements of large-scale cosmic microwave background (CMB) E-mode polarization from the Cosmology Large Angular Scale Surveyor (CLASS) 90 GHz data. Using 115 det-yr of observations collected through 2024 with a variable-delay polarization modulator, we achieved a polarization sensitivity of $78\,\mathrm{μK\,arcmin}$, comparable to Planck at similar frequencies (100 and 143 GHz). The analysis demonstrates effective mitigation of systematic errors and addresses challenges to large-angular-scale power recovery posed by time-domain filtering in maximum-likelihood map-making. A novel implementation of the pixel-space transfer matrix is introduced, which enables efficient filtering simulations and bias correction in the power spectrum using the quadratic cross-spectrum estimator. Overall, we achieved an unbiased time-domain filtering correction to recover the largest angular scale polarization, with the only power deficit, arising from map-making non-linearity, being characterized as less than $3\%$. Through cross-correlation with Planck, we detected the cosmic reionization at $99.4\%$ significance and measured the reionization optical depth $τ=0.053^{+0.018}_{-0.019}$, marking the first ground-based attempt at such a measurement. At intermediate angular scales ($\ell>30$), our results, both independently and in cross-correlation with Planck, remain fully consistent with Planck's measurements.

astro-ph.CO↗

arXiv2025

VIDEX: A Disaggregated and Extensible Virtual Index for the Cloud and AI Era

Virtual index, also known as hypothetical indexes, play a crucial role in database query optimization. However, with the rapid advancement of cloud computing and AI-driven models for database optimization, traditional virtual index approaches face significant challenges. Cloud-native environments often prohibit direct conducting query optimization process on production databases due to stability requirements and data privacy concerns. Moreover, while AI models show promising progress, their integration with database systems poses challenges in system complexity, inference acceleration, and model hot updates. In this paper, we present VIDEX, a three-layer disaggregated architecture that decouples database instances, the virtual index optimizer, and algorithm services, providing standardized interfaces for AI model integration. Users can configure VIDEX by either collecting production statistics or by loading from a prepared file; this setup allows for high-accurate what-if analyses based on virtual indexes, achieving query plans that are identical to those of the production instance. Additionally, users can freely integrate new AI-driven algorithms into VIDEX. VIDEX has been successfully deployed at ByteDance, serving thousands of MySQL instances daily and over millions of SQL queries for index optimization tasks.

cs.DB↗

arXiv2025

ChatTS: Aligning Time Series with LLMs via Synthetic Data for Enhanced Understanding and Reasoning

Understanding time series is crucial for its application in real-world scenarios. Recently, large language models (LLMs) have been increasingly applied to time series tasks, leveraging their strong language capabilities to enhance various applications. However, research on multimodal LLMs (MLLMs) for time series understanding and reasoning remains limited, primarily due to the scarcity of high-quality datasets that align time series with textual information. This paper introduces ChatTS, a novel MLLM designed for time series analysis. ChatTS treats time series as a modality, similar to how vision MLLMs process images, enabling it to perform both understanding and reasoning with time series. To address the scarcity of training data, we propose an attribute-based method for generating synthetic time series with detailed attribute descriptions. We further introduce Time Series Evol-Instruct, a novel approach that generates diverse time series Q&As, enhancing the model's reasoning capabilities. To the best of our knowledge, ChatTS is the first TS-MLLM that takes multivariate time series as input for understanding and reasoning, which is fine-tuned exclusively on synthetic datasets. We evaluate its performance using benchmark datasets with real-world data, including six alignment tasks and four reasoning tasks. Our results show that ChatTS significantly outperforms existing vision-based MLLMs (e.g., GPT-4o) and text/agent-based LLMs, achieving a 46.0% improvement in alignment tasks and a 25.8% improvement in reasoning tasks. We have open-sourced the source code, model checkpoint and datasets at https://github.com/NetManAIOps/ChatTS.

cs.AI↗

arXiv2025

TickIt: Leveraging Large Language Models for Automated Ticket Escalation

In large-scale cloud service systems, support tickets serve as a critical mechanism for resolving customer issues and maintaining service quality. However, traditional manual ticket escalation processes encounter significant challenges, including inefficiency, inaccuracy, and difficulty in handling the high volume and complexity of tickets. While previous research has proposed various machine learning models for ticket classification, these approaches often overlook the practical demands of real-world escalations, such as dynamic ticket updates, topic-specific routing, and the analysis of ticket relationships. To bridge this gap, this paper introduces TickIt, an innovative online ticket escalation framework powered by Large Language Models. TickIt enables topic-aware, dynamic, and relationship-driven ticket escalations by continuously updating ticket states, assigning tickets to the most appropriate support teams, exploring ticket correlations, and leveraging category-guided supervised fine-tuning to continuously improve its performance. By deploying TickIt in ByteDance's cloud service platform Volcano Engine, we validate its efficacy and practicality, marking a significant advancement in the field of automated ticket escalation for large-scale cloud service systems.

cs.SE↗

arXiv2025

InstantSticker: Realistic Decal Blending via Disentangled Object Reconstruction

We present InstantSticker, a disentangled reconstruction pipeline based on Image-Based Lighting (IBL), which focuses on highly realistic decal blending, simulates stickers attached to the reconstructed surface, and allows for instant editing and real-time rendering. To achieve stereoscopic impression of the decal, we introduce shadow factor into IBL, which can be adaptively optimized during training. This allows the shadow brightness of surfaces to be accurately decomposed rather than baked into the diffuse color, ensuring that the edited texture exhibits authentic shading. To address the issues of warping and blurriness in previous methods, we apply As-Rigid-As-Possible (ARAP) parameterization to pre-unfold a specified area of the mesh and use the local UV mapping combined with a neural texture map to enhance the ability to express high-frequency details in that area. For instant editing, we utilize the Disney BRDF model, explicitly defining material colors with 3-channel diffuse albedo. This enables instant replacement of albedo RGB values during the editing process, avoiding the prolonged optimization required in previous approaches. In our experiment, we introduce the Ratio Variance Warping (RVW) metric to evaluate the local geometric warping of the decal area. Extensive experimental results demonstrate that our method surpasses previous decal blending methods in terms of editing quality, editing speed and rendering speed, achieving the state-of-the-art.

cs.CV↗

arXiv2025

Advancing the Economic and Environmental Sustainability of Rare Earth Element Recovery from Phosphogypsum

Transitioning to green energy technologies requires more sustainable and secure rare earth elements (REE) production. The current production of rare earth oxides (REOs) is completed by an energy and chemically intensive process from the mining of REE ores. Investigations into a more sustainable supply of REEs from secondary sources, such as toxic phosphogypsum (PG) waste, is vital to securing the REE supply chain. However, conventional solvent extraction to recover dilute REEs from PG waste is inefficient and has high environmental impact. In this work, we propose a treatment train for the recovery of REEs from PG which includes a bio-inspired adsorptive separation to generate a stream of pure REEs, and we assess its financial viability and environmental impacts under uncertainties through a "probabilistic sustainability" framework integrating life cycle assessment (LCA) and techno-economic analysis (TEA). Results show that in 87% of baseline scenario simulations, the internal rate of return (IRR) exceeded 15%, indicating that this system has the potential to be profitable. However, environmental impacts of the system are mixed. Specifically, the proposed system outperforms conventional systems in ecosystem quality and resource depletion, but has higher human health impacts. Scenario analysis shows that the system is profitable at capacities larger than 100,000 kg*hr-1*PG for PG with REE content above 0.5 wt%. The most dilute PG sources (0.02-0.1 wt% REE) are inaccessible using the current process scheme (limited by the cost of acid and subsequent neutralization) requiring further examination of new process schemes and improvements in technological performance. Overall, this study evaluates the sustainability of a first-of-its-kind REE recovery process from PG and uses these results to provide clear direction for advancing sustainable REE recovery from secondary sources.

physics.soc-ph↗

arXiv2025

PLM4NDV: Minimizing Data Access for Number of Distinct Values Estimation with Pre-trained Language Models

Number of Distinct Values (NDV) estimation of a multiset/column is a basis for many data management tasks, especially within databases. Despite decades of research, most existing methods require either a significant amount of samples through uniform random sampling or access to the entire column to produce estimates, leading to substantial data access costs and potentially ineffective estimations in scenarios with limited data access. In this paper, we propose leveraging semantic information, i.e., schema, to address these challenges. The schema contains rich semantic information that can benefit the NDV estimation. To this end, we propose PLM4NDV, a learned method incorporating Pre-trained Language Models (PLMs) to extract semantic schema information for NDV estimation. Specifically, PLM4NDV leverages the semantics of the target column and the corresponding table to gain a comprehensive understanding of the column's meaning. By using the semantics, PLM4NDV reduces data access costs, provides accurate NDV estimation, and can even operate effectively without any data access. Extensive experiments on a large-scale real-world dataset demonstrate the superiority of PLM4NDV over baseline methods. Our code is available at https://github.com/bytedance/plm4ndv.

cs.DB↗

arXiv2025

SPRING: Systematic Profiling of Randomly Interconnected Neural Networks Generated by HLS

Profiling is important for performance optimization by providing real-time observations and measurements of important parameters of hardware execution. Existing profiling tools for High-Level Synthesis (HLS) IPs running on FPGAs are far less mature compared with those developed for fixed CPU and GPU architectures and they still lag behind mainly due to their dynamic architecture. This limitation is reflected in the typical approach of extracting monitoring signals off of an FPGA device individually from dedicated ports, using one BRAM per signal for temporary information storage, or embedding vendor specific primitives to manually analyze the waveform. In this paper, we propose a systematic profiling method tailored to the dynamic nature of FPGA systems, particularly suitable for streaming accelerators. Instead of relying on signal extraction, the proposed profiling stream flows alongside the actual data, dynamically splitting and merging in synchrony with the data stream, and is ultimately directed to the processing system (PS) side. We conducted a preliminary evaluation of this method on randomly interconnected neural networks (RINNs) using the FIFO fullness metric, with co-simulation results for validation.

cs.AR↗

arXiv2025

RFUAV: A Benchmark Dataset for Unmanned Aerial Vehicle Detection and Identification

In this paper, we propose RFUAV as a new benchmark dataset for radio-frequency based (RF-based) unmanned aerial vehicle (UAV) identification and address the following challenges: Firstly, many existing datasets feature a restricted variety of drone types and insufficient volumes of raw data, which fail to meet the demands of practical applications. Secondly, existing datasets often lack raw data covering a broad range of signal-to-noise ratios (SNR), or do not provide tools for transforming raw data to different SNR levels. This limitation undermines the validity of model training and evaluation. Lastly, many existing datasets do not offer open-access evaluation tools, leading to a lack of unified evaluation standards in current research within this field. RFUAV comprises approximately 1.3 TB of raw frequency data collected from 37 distinct UAVs using the Universal Software Radio Peripheral (USRP) device in real-world environments. Through in-depth analysis of the RF data in RFUAV, we define a drone feature sequence called RF drone fingerprint, which aids in distinguishing drone signals. In addition to the dataset, RFUAV provides a baseline preprocessing method and model evaluation tools. Rigorous experiments demonstrate that these preprocessing methods achieve state-of-the-art (SOTA) performance using the provided evaluation tools. The RFUAV dataset and baseline implementation are publicly available at https://github.com/kitoweeknd/RFUAV/.

cs.RO↗

arXiv2025

ZuantuSet: A Collection of Historical Chinese Visualizations and Illustrations

Historical visualizations are a valuable resource for studying the history of visualization and inspecting the cultural context where they were created. When investigating historical visualizations, it is essential to consider contributions from different cultural frameworks to gain a comprehensive understanding. While there is extensive research on historical visualizations within the European cultural framework, this work shifts the focus to ancient China, a cultural context that remains underexplored by visualization researchers. To this aim, we propose a semi-automatic pipeline to collect, extract, and label historical Chinese visualizations. Through the pipeline, we curate ZuantuSet, a dataset with over 71K visualizations and 108K illustrations. We analyze distinctive design patterns of historical Chinese visualizations and their potential causes within the context of Chinese history and culture. We illustrate potential usage scenarios for this dataset, summarize the unique challenges and solutions associated with collecting historical Chinese visualizations, and outline future research directions.

cs.HC↗

arXiv2025

AIBrix: Towards Scalable, Cost-Effective Large Language Model Inference Infrastructure

We introduce AIBrix, a cloud-native, open-source framework designed to optimize and simplify large-scale LLM deployment in cloud environments. Unlike traditional cloud-native stacks, AIBrix follows a co-design philosophy, ensuring every layer of the infrastructure is purpose-built for seamless integration with inference engines like vLLM. AIBrix introduces several key innovations to reduce inference costs and enhance performance including high-density LoRA management for dynamic adapter scheduling, LLM-specific autoscalers, and prefix-aware, load-aware routing. To further improve efficiency, AIBrix incorporates a distributed KV cache, boosting token reuse across nodes, leading to a 50% increase in throughput and a 70% reduction in inference latency. AIBrix also supports unified AI runtime which streamlines model management while maintaining vendor-agnostic engine compatibility. For large-scale multi-node inference, AIBrix employs hybrid orchestration -- leveraging Kubernetes for coarse-grained scheduling and Ray for fine-grained execution -- to balance efficiency and flexibility. Additionally, an SLO-driven GPU optimizer dynamically adjusts resource allocations, optimizing heterogeneous serving to maximize cost efficiency while maintaining service guarantees. Finally, AIBrix enhances system reliability with AI accelerator diagnostic tools, enabling automated failure detection and mock-up testing to improve fault resilience. AIBrix is available at https://github.com/vllm-project/aibrix.

cs.DC↗

arXiv2025

TSEML: A task-specific embedding-based method for few-shot classification of cancer molecular subtypes

Molecular subtyping of cancer is recognized as a critical and challenging upstream task for personalized therapy. Existing deep learning methods have achieved significant performance in this domain when abundant data samples are available. However, the acquisition of densely labeled samples for cancer molecular subtypes remains a significant challenge for conventional data-intensive deep learning approaches. In this work, we focus on the few-shot molecular subtype prediction problem in heterogeneous and small cancer datasets, aiming to enhance precise diagnosis and personalized treatment. We first construct a new few-shot dataset for cancer molecular subtype classification and auxiliary cancer classification, named TCGA Few-Shot, from existing publicly available datasets. To effectively leverage the relevant knowledge from both tasks, we introduce a task-specific embedding-based meta-learning framework (TSEML). TSEML leverages the synergistic strengths of a model-agnostic meta-learning (MAML) approach and a prototypical network (ProtoNet) to capture diverse and fine-grained features. Comparative experiments conducted on the TCGA Few-Shot dataset demonstrate that our TSEML framework achieves superior performance in addressing the problem of few-shot molecular subtype classification.

q-bio.QM↗

arXiv2024

Unveiling hole-facilitated amorphisation in pressure-induced phase transformation of silicon

Pressure-induced phase transformation occurs during silicon (Si) wafering processes. \b{eta}-tin (Si-II) phase is formed at high pressures, followed by the transformation to Si-XII, Si-III or/and amorphous Si (α-Si) phases during the subsequent decompression. While the imposed pressure and its release rate are known to dictate the phase transformation of Si, the effect of charge carriers are ignored. Here, we experimentally unveil that the increased hole concentration facilitates the amorphization in the pressure-induced phase transformation of Si. The underlying mechanism is elucidated by the theoretical calculations based on machine-learning interatomic potentials. The hole-facilitated amorphization is also experimentally confirmed to occur in the indented Ge, GaAs or SiC. We discover that hole concentration is another determining factor for the pressure-induced phase transformations of the industrially important semiconductors.

cond-mat.mtrl-sci↗

arXiv2024

Outlier Detection with Cluster Catch Digraphs

This paper introduces a novel family of outlier detection algorithms based on Cluster Catch Digraphs (CCDs), specifically tailored to address the challenges of high dimensionality and varying cluster shapes, which deteriorate the performance of most traditional outlier detection methods. We propose the Uniformity-Based CCD with Mutual Catch Graph (U-MCCD), the Uniformity- and Neighbor-Based CCD with Mutual Catch Graph (UN-MCCD), and their shape-adaptive variants (SU-MCCD and SUN-MCCD), which are designed to detect outliers in data sets with arbitrary cluster shapes and high dimensions. We present the advantages and shortcomings of these algorithms and provide the motivation or need to define each particular algorithm. Through comprehensive Monte Carlo simulations, we assess their performance and demonstrate the robustness and effectiveness of our algorithms across various settings and contamination levels. We also illustrate the use of our algorithms on various real-life data sets. The U-MCCD algorithm efficiently identifies outliers while maintaining high true negative rates, and the SU-MCCD algorithm shows substantial improvement in handling non-uniform clusters. Additionally, the UN-MCCD and SUN-MCCD algorithms address the limitations of existing methods in high-dimensional spaces by utilizing Nearest Neighbor Distances (NND) for clustering and outlier detection. Our results indicate that these novel algorithms offer substantial advancements in the accuracy and adaptability of outlier detection, providing a valuable tool for various real-world applications. Keyword: Outlier detection, Graph-based clustering, Cluster catch digraphs, $k$-nearest-neighborhood, Mutual catch graphs, Nearest neighbor distance.

stat.ML↗

arXiv2024

Toward Tiny and High-quality Facial Makeup with Data Amplify Learning

Contemporary makeup approaches primarily hinge on unpaired learning paradigms, yet they grapple with the challenges of inaccurate supervision (e.g., face misalignment) and sophisticated facial prompts (including face parsing, and landmark detection). These challenges prohibit low-cost deployment of facial makeup models, especially on mobile devices. To solve above problems, we propose a brand-new learning paradigm, termed "Data Amplify Learning (DAL)," alongside a compact makeup model named "TinyBeauty." The core idea of DAL lies in employing a Diffusion-based Data Amplifier (DDA) to "amplify" limited images for the model training, thereby enabling accurate pixel-to-pixel supervision with merely a handful of annotations. Two pivotal innovations in DDA facilitate the above training approach: (1) A Residual Diffusion Model (RDM) is designed to generate high-fidelity detail and circumvent the detail vanishing problem in the vanilla diffusion models; (2) A Fine-Grained Makeup Module (FGMM) is proposed to achieve precise makeup control and combination while retaining face identity. Coupled with DAL, TinyBeauty necessitates merely 80K parameters to achieve a state-of-the-art performance without intricate face prompts. Meanwhile, TinyBeauty achieves a remarkable inference speed of up to 460 fps on the iPhone 13. Extensive experiments show that DAL can produce highly competitive makeup models using only 5 image pairs.

cs.CV↗