arXiv subjects

Zhiying Wu

Publications and source records attributed to Zhiying Wu.

14 recordsLinked to original sources

arXiv2026

FlashPrefill V2: Block-Sparse Prefill Attention for Long-Context LLM Serving

Long-context modeling is a pivotal capability for Large Language Models, yet the quadratic complexity of attention remains a critical bottleneck, particularly during the compute-intensive prefilling phase. Our previous work, FlashPrefill, mitigates this cost through instantaneous pattern discovery and max-based dynamic thresholding; however, it remains an algorithmic prototype that is still distant from production deployment. In this paper, we present FlashPrefill V2, which evolves FlashPrefill from a prototype toward practical long-context serving along three dimensions. First, we introduce a mean correction term that effectively suppresses the approximation error, keeping performance degradation manageable even at extreme sparsity levels. Second, we redesign the sparse attention operator with PackGQA memory access, warp specialization, and pingpong pipelining, fully aligning with the latest FlashAttention-3/4 implementations and supporting FP8 inference to meet practical quantization requirements. Third, FlashPrefill V2 natively supports paged KV cache and continuous batching, allowing integration as an attention backend in modern inference frameworks such as SGLang. Extensive evaluations on NVIDIA H20 GPUs---among the most widely deployed inference accelerators---demonstrate that FlashPrefill V2 delivers up to 47.26x and 27.19x speedups over FlashAttention-2 at 128K context length under FP8 and BF16 precision, respectively, and, in FP8, still achieves a 30.49x speedup against an FA3/4-aligned dense baseline.

cs.CL↗

arXiv2026

CelerLog: Fast Log Parsing via Dynamic Routing

Log parsing is a fundamental step for automated log analysis, which transforms raw log messages into structured formats. Existing syntax-based parsers struggle with complex logs because they lack semantic reasoning ability. Emerging LLM-powered semantic parsers achieve high accuracy but suffer from prohibitive latency and token costs because they apply semantic inference across all logs. Our key observation is that not all logs necessitate complex semantic understanding: a vast majority of logs exhibit repetitive patterns that can be extracted via straightforward statistical analysis. Driven by this insight, we propose CelerLog, a fast and effective log parser. CelerLog introduces a dynamic routing mechanism to classify logs into dense and sparse groups. Logs with strong statistical patterns (dense groups) are processed by an efficient statistical processor, whereas the sparse groups lacking such patterns are routed to an LLM for semantic inference. This hybrid strategy avoids unnecessary LLM invocations. Extensive experiments on 14 public datasets show that CelerLog achieves leading performance over state-of-the-art baselines and is 7.9x to 18.6x faster than LLM methods and up to 1.5x faster than Drain. Additionally, it reduces costs by decreasing token consumption by 80.2% - 94.1% and LLM invocations by 86.4% - 90.9%.

cs.SE↗

arXiv2026

UniPrefill: Universal Long-Context Prefill Acceleration via Block-wise Dynamic Sparsification

As large language models (LLMs) continue to advance rapidly, they are becoming increasingly capable while simultaneously demanding ever-longer context lengths. To improve the inference efficiency of long-context processing, several novel low-complexity hybrid architectures have recently been proposed, effectively alleviating the computational burden of long-context inference. However, existing research on long-context prefill acceleration remains predominantly focused on sparse attention mechanisms, which achieve their maximum speedup only on full-attention models. When transferred to emerging architectures--such as linear/full attention hybrids or sliding window/full attention hybrids--these prefill acceleration approaches suffer significant performance degradation. Furthermore, such methods are generally incompatible with continuous batching, making them difficult to integrate into modern inference engines such as vLLM. To this end, we propose UniPrefill, a prefill acceleration framework applicable to virtually any model architecture, which directly accelerates the model's computation at the token level. We further implement UniPrefill as a continuous batching operator and extend vLLM's scheduling strategy to natively support prefill-decode co-processing and tensor parallel for UniPrefill, enabling its seamless integration into vLLM. UniPrefill achieves up to 2.1x speedup in Time-To-First-Token (TTFT), with the acceleration becoming increasingly pronounced as the number of concurrent requests grows.

cs.CL↗

arXiv2026

FlashPrefill: Instantaneous Pattern Discovery and Thresholding for Ultra-Fast Long-Context Prefilling

Long-context modeling is a pivotal capability for Large Language Models, yet the quadratic complexity of attention remains a critical bottleneck, particularly during the compute-intensive prefilling phase. While various sparse attention mechanisms have been explored, they typically suffer from either significant search latency or insufficient sparsity. In this paper, we propose FlashPrefill, a framework enabling ultra-fast prefilling via instantaneous pattern discovery and thresholding. FlashPrefill leverages a fast block-searching technique to simultaneously locate dynamic vertical, slash, and block-sparse attention patterns. Crucially, it introduces a dynamic thresholding mechanism that bypasses the prohibitive overhead of sorting or accumulating attention scores while effectively eliminating the long-tail distribution to enhance sparsity. Extensive evaluations demonstrate that FlashPrefill achieves a substantial leap in efficiency, delivering an unprecedented 27.78x speedup on 256K sequences. Notably, unlike existing methods that incur efficiency degradation on shorter contexts, FlashPrefill maintains a 1.71x speedup even at a 4K context length, demonstrating its robustness and practical utility across varying sequence scales.

cs.CL↗

arXiv2025

F2PASeg: Feature Fusion for Pituitary Anatomy Segmentation in Endoscopic Surgery

Pituitary tumors often cause deformation or encapsulation of adjacent vital structures. Anatomical structure segmentation can provide surgeons with early warnings of regions that pose surgical risks, thereby enhancing the safety of pituitary surgery. However, pixel-level annotated video stream datasets for pituitary surgeries are extremely rare. To address this challenge, we introduce a new dataset for Pituitary Anatomy Segmentation (PAS). PAS comprises 7,845 time-coherent images extracted from 120 videos. To mitigate class imbalance, we apply data augmentation techniques that simulate the presence of surgical instruments in the training data. One major challenge in pituitary anatomy segmentation is the inconsistency in feature representation due to occlusions, camera motion, and surgical bleeding. By incorporating a Feature Fusion module, F2PASeg is proposed to refine anatomical structure segmentation by leveraging both high-resolution image features and deep semantic embeddings, enhancing robustness against intraoperative variations. Experimental results demonstrate that F2PASeg consistently segments critical anatomical structures in real time, providing a reliable solution for intraoperative pituitary surgery planning. Code: https://github.com/paulili08/F2PASeg.

cs.CV↗

arXiv2025

MM2CT: MR-to-CT translation for multi-modal image fusion with mamba

Magnetic resonance (MR)-to-computed tomography (CT) translation offers significant advantages, including the elimination of radiation exposure associated with CT scans and the mitigation of imaging artifacts caused by patient motion. The existing approaches are based on single-modality MR-to-CT translation, with limited research exploring multimodal fusion. To address this limitation, we introduce Multi-modal MR to CT (MM2CT) translation method by leveraging multimodal T1- and T2-weighted MRI data, an innovative Mamba-based framework for multi-modal medical image synthesis. Mamba effectively overcomes the limited local receptive field in CNNs and the high computational complexity issues in Transformers. MM2CT leverages this advantage to maintain long-range dependencies modeling capabilities while achieving multi-modal MR feature integration. Additionally, we incorporate a dynamic local convolution module and a dynamic enhancement module to improve MRI-to-CT synthesis. The experiments on a public pelvis dataset demonstrate that MM2CT achieves state-of-the-art performance in terms of Structural Similarity Index Measure (SSIM) and Peak Signal-to-Noise Ratio (PSNR). Our code is publicly available at https://github.com/Gots-ch/MM2CT.

eess.IV↗

arXiv2025

SolPhishHunter: Towards Detecting and Understanding Phishing on Solana

Solana is a rapidly evolving blockchain platform that has attracted an increasing number of users. However, this growth has also drawn the attention of malicious actors, with some phishers extending their reach into the Solana ecosystem. Unlike platforms such as Ethereum, Solana has distinct designs of accounts and transactions, leading to the emergence of new types of phishing transactions that we term SolPhish. We define three types of SolPhish and develop a detection tool called SolPhishHunter. Utilizing SolPhishHunter, we detect a total of 8,058 instances of SolPhish and conduct an empirical analysis of these detected cases. Our analysis explores the distribution and impact of SolPhish, the characteristics of the phishers, and the relationships among phishing gangs. Particularly, the detected SolPhish transactions have resulted in nearly \$1.1 million in losses for victims. We report our detection results to the community and construct SolPhishDataset, the \emph{first} Solana phishing-related dataset in academia.

cs.CR↗

arXiv2025

Baichuan 2: Open Large-scale Language Models

Large language models (LLMs) have demonstrated remarkable performance on a variety of natural language tasks based on just a few examples of natural language instructions, reducing the need for extensive feature engineering. However, most powerful LLMs are closed-source or limited in their capability for languages other than English. In this technical report, we present Baichuan 2, a series of large-scale multilingual language models containing 7 billion and 13 billion parameters, trained from scratch, on 2.6 trillion tokens. Baichuan 2 matches or outperforms other open-source models of similar size on public benchmarks like MMLU, CMMLU, GSM8K, and HumanEval. Furthermore, Baichuan 2 excels in vertical domains such as medicine and law. We will release all pre-training model checkpoints to benefit the research community in better understanding the training dynamics of Baichuan 2.

cs.CL↗

arXiv2025

Baichuan-M1: Pushing the Medical Capability of Large Language Models

The current generation of large language models (LLMs) is typically designed for broad, general-purpose applications, while domain-specific LLMs, especially in vertical fields like medicine, remain relatively scarce. In particular, the development of highly efficient and practical LLMs for the medical domain is challenging due to the complexity of medical knowledge and the limited availability of high-quality data. To bridge this gap, we introduce Baichuan-M1, a series of large language models specifically optimized for medical applications. Unlike traditional approaches that simply continue pretraining on existing models or apply post-training to a general base model, Baichuan-M1 is trained from scratch with a dedicated focus on enhancing medical capabilities. Our model is trained on 20 trillion tokens and incorporates a range of effective training methods that strike a balance between general capabilities and medical expertise. As a result, Baichuan-M1 not only performs strongly across general domains such as mathematics and coding but also excels in specialized medical fields. We have open-sourced Baichuan-M1-14B, a mini version of our model, which can be accessed through the following links.

cs.CL↗

arXiv2025

Safeguarding Blockchain Ecosystem: Understanding and Detecting Attack Transactions on Cross-chain Bridges

Cross-chain bridges are essential decentralized applications (DApps) to facilitate interoperability between different blockchain networks. Unlike regular DApps, the functionality of cross-chain bridges relies on the collaboration of information both on and off the chain, which exposes them to a wider risk of attacks. According to our statistics, attacks on cross-chain bridges have resulted in losses of nearly 4.3 billion dollars since 2021. Therefore, it is particularly necessary to understand and detect attacks on cross-chain bridges. In this paper, we collect the largest number of cross-chain bridge attack incidents to date, including 49 attacks that occurred between June 2021 and September 2024. Our analysis reveal that attacks against cross-chain business logic cause significantly more damage than those that do not. These cross-chain attacks exhibit different patterns compared to normal transactions in terms of call structure, which effectively indicates potential attack behaviors. Given the significant losses in these cases and the scarcity of related research, this paper aims to detect attacks against cross-chain business logic, and propose the BridgeGuard tool. Specifically, BridgeGuard models cross-chain transactions from a graph perspective, and employs a two-stage detection framework comprising global and local graph mining to identify attack patterns in cross-chain transactions. We conduct multiple experiments on the datasets with 203 attack transactions and 40,000 normal cross-chain transactions. The results show that BridgeGuard's reported recall score is 36.32\% higher than that of state-of-the-art tools and can detect unknown attack transactions.

cs.CR↗

arXiv2025

Baichuan-Omni-1.5 Technical Report

We introduce Baichuan-Omni-1.5, an omni-modal model that not only has omni-modal understanding capabilities but also provides end-to-end audio generation capabilities. To achieve fluent and high-quality interaction across modalities without compromising the capabilities of any modality, we prioritized optimizing three key aspects. First, we establish a comprehensive data cleaning and synthesis pipeline for multimodal data, obtaining about 500B high-quality data (text, audio, and vision). Second, an audio-tokenizer (Baichuan-Audio-Tokenizer) has been designed to capture both semantic and acoustic information from audio, enabling seamless integration and enhanced compatibility with MLLM. Lastly, we designed a multi-stage training strategy that progressively integrates multimodal alignment and multitask fine-tuning, ensuring effective synergy across all modalities. Baichuan-Omni-1.5 leads contemporary models (including GPT4o-mini and MiniCPM-o 2.6) in terms of comprehensive omni-modal capabilities. Notably, it achieves results comparable to leading models such as Qwen2-VL-72B across various multimodal medical benchmarks.

cs.CL↗

arXiv2024

INT-FlashAttention: Enabling Flash Attention for INT8 Quantization

As the foundation of large language models (LLMs), self-attention module faces the challenge of quadratic time and memory complexity with respect to sequence length. FlashAttention accelerates attention computation and reduces its memory usage by leveraging the GPU memory hierarchy. A promising research direction is to integrate FlashAttention with quantization methods. This paper introduces INT-FlashAttention, the first INT8 quantization architecture compatible with the forward workflow of FlashAttention, which significantly improves the inference speed of FlashAttention on Ampere GPUs. We implement our INT-FlashAttention prototype with fully INT8 activations and general matrix-multiplication (GEMM) kernels, making it the first attention operator with fully INT8 input. As a general token-level post-training quantization framework, INT-FlashAttention is also compatible with other data formats like INT4, etc. Experimental results show INT-FlashAttention achieves 72% faster inference speed and 82% smaller quantization error compared to standard FlashAttention with FP16 and FP8 data format.

cs.LG↗

arXiv2023

Bubble or Not: Measurements, Analyses, and Findings on the Ethereum ERC721 and ERC1155 Non-fungible Token Ecosystem

The non-fungible token (NFT) is an emergent type of cryptocurrency that has garnered extensive attention since its inception. The uniqueness, indivisibility and humanistic value of NFTs are the key characteristics that distinguish them from traditional tokens. The market capitalization of NFT reached 21.5 billion USD in 2021, almost 200 times of all previous transactions. However, the subsequent rapid decline in NFT market fever in the second quarter of 2022 casts doubts on the ostensible boom in the NFT market. To date, there has been no comprehensive and systematic study of the NFT trade market or of the NFT bubble and hype phenomenon. To fill this gap, we conduct an in-depth investigation of the whole Ethereum ERC721 and ERC1155 NFT ecosystem via graph analysis and apply several metrics to measure the characteristics of NFTs. By collecting data from the whole blockchain, we construct three graphs, namely NFT create graph, NFT transfer graph, and NFT hold graph, to characterize the NFT traders, analyze the characteristics of NFTs, and discover many observations and insights. Moreover, we propose new indicators to quantify the activeness and value of NFT and propose an algorithm that combines indicators and graph analyses to find bubble NFTs. Real-world cases demonstrate that our indicators and approach can be used to discern bubble NFTs effectively.

cs.CY↗

arXiv2022

TRacer: Scalable Graph-based Transaction Tracing for Account-based Blockchain Trading Systems

Security incidents such as scams and hacks, have become a major threat to the health of the blockchain ecosystem, causing billions of dollars in losses each year for blockchain users. To reveal the real-world entities behind the pseudonymous blockchain account and recover the stolen funds from the massive transaction data, much effort has been devoted to tracing the flow of illicit funds in blockchains recently. However, most current tracing approaches based on heuristics and taint analysis have limitations in terms of universality, effectiveness, and efficiency. This paper models the blockchain transaction records as a blockchain transaction graph and tackles blockchain transaction tracing as a graph searching task. We propose TRacer, a scalable transaction tracing tool for account-based blockchains. To infer the relevance between accounts during graph searching, we develop a novel personalized PageRank method in TRacer based on the directed, weighted, temporal, and multi-relationship blockchain transaction graphs. To the best of our knowledge, TRacer is the first intelligent transaction tracing tool in account-based blockchains that can handle complex transaction actions in decentralized finance (DeFi). Experimental results and theoretical analysis prove that TRacer can complete the transaction tracing task effectively at a low cost. All codes of TRacer are available at GitHub.

cs.CR↗