arXiv subjects

Chenyuan Wu

Publications and source records attributed to Chenyuan Wu.

10 recordsLinked to original sources

arXiv2026

AutoPilot: Learning to Steer High Speed Robust BFT

Recent Byzantine Fault Tolerant (BFT) protocols achieve strong performance by combining the low-latency advantages of leader-based BFT protocols with the high-throughput benefits of DAG-based data dissemination. Despite exposing a wide spectrum of internal tunable parameters, these protocols typically rely on static and heuristic configurations, which leads to performance degradation under dynamic workloads, heterogeneous network conditions, and evolving adversarial behaviors. In this paper, we present AutoPilot, a reinforcement learning-based framework that continuously monitors runtime conditions and dynamically adjusts protocol parameters online to optimize consensus performance. To ensure robustness, AutoPilot coordinates learning in a decentralized manner, providing resilience against adversarial data pollution. We implement AutoPilot on top of Autobahn, a state-of-the-art, highspeed, robust BFT protocol, and evaluate it across diverse dynamic environments. Experimental results demonstrate that AutoPilot quickly converges to the optimal configuration under changing environments, reduces end-to-end latency by 49.8% compared to the default protocol configuration, and outperforms random configuration exploration by 73.3%.

cs.DC↗

arXiv2026

OmniGen2: Towards Instruction-Aligned Multimodal Generation

In this work, we introduce OmniGen2, a versatile and open-source generative model designed to provide a unified solution for diverse generation tasks, including text-to-image, image editing, and in-context generation. Unlike OmniGen v1, OmniGen2 features two distinct decoding pathways for text and image modalities, utilizing unshared parameters and a decoupled image tokenizer. This design enables OmniGen2 to build upon existing multimodal understanding models without the need to re-adapt VAE inputs, thereby preserving the original text generation capabilities. To facilitate the training of OmniGen2, we developed comprehensive data construction pipelines, encompassing image editing and in-context generation data. Additionally, we introduce a reflection mechanism tailored for image generation tasks and curate a dedicated reflection dataset based on OmniGen2. Despite its relatively modest parameter size, OmniGen2 achieves competitive results on multiple task benchmarks, including text-to-image and image editing. To further evaluate in-context generation, also referred to as subject-driven tasks, we introduce a new benchmark named OmniContext. OmniGen2 achieves state-of-the-art performance among open-source models in terms of consistency. We will release our models, training code, datasets, and data construction pipeline to support future research in this field. Project Page: https://vectorspacelab.github.io/OmniGen2; GitHub Link: https://github.com/VectorSpaceLab/OmniGen2

cs.CV↗

arXiv2026

EditScore: Unlocking Online RL for Image Editing via High-Fidelity Reward Modeling

Instruction-guided image editing has achieved remarkable progress, yet current models still face challenges with complex instructions and often require multiple samples to produce a desired result. Reinforcement Learning (RL) offers a promising solution, but its adoption in image editing has been severely hindered by the lack of a high-fidelity, efficient reward signal. In this work, we present a comprehensive methodology to overcome this barrier, centered on the development of a state-of-the-art, specialized reward model. We first introduce EditReward-Bench, a comprehensive benchmark to systematically evaluate reward models on editing quality. Building on this benchmark, we develop EditScore, a series of reward models (7B-72B) for evaluating the quality of instruction-guided image editing. Through meticulous data curation and filtering, EditScore effectively matches the performance of learning proprietary VLMs. Furthermore, coupled with an effective self-ensemble strategy tailored for the generative nature of EditScore, our largest variant even surpasses GPT-5 in the benchmark. We then demonstrate that a high-fidelity reward model is the key to unlocking online RL for image editing. Our experiments show that, while even the largest open-source VLMs fail to provide an effective learning signal, EditScore enables efficient and robust policy optimization. Applying our framework to a strong base model, OmniGen2, results in a final model that shows a substantial and consistent performance uplift. Overall, this work provides the first systematic path from benchmarking to reward modeling to RL training in image editing, showing that a high-fidelity, domain-specialized reward model is the key to unlocking the full potential of RL in this domain.

cs.CV↗

arXiv2024

Lighter And Better: Towards Flexible Context Adaptation For Retrieval Augmented Generation

The existing Retrieval-Augmented Generation (RAG) systems face significant challenges in terms of cost and effectiveness. On one hand, they need to encode the lengthy retrieved contexts before responding to the input tasks, which imposes substantial computational overhead. On the other hand, directly using generic Large Language Models (LLMs) often leads to sub-optimal answers, while task-specific fine-tuning may compromise the LLMs' general capabilities. To address these challenges, we introduce a novel approach called FlexRAG (Flexible Context Adaptation for RAG). In this approach, the retrieved contexts are compressed into compact embeddings before being encoded by the LLMs. Simultaneously, these compressed embeddings are optimized to enhance downstream RAG performance. A key feature of FlexRAG is its flexibility, which enables effective support for diverse compression ratios and selective preservation of important contexts. Thanks to these technical designs, FlexRAG achieves superior generation quality while significantly reducing running costs. Comprehensive experiments on various question-answering datasets validate our approach as a cost-effective and flexible solution for RAG systems.

cs.CL↗

arXiv2024

BFTBrain: Adaptive BFT Consensus with Reinforcement Learning

This paper presents BFTBrain, a reinforcement learning (RL) based Byzantine fault-tolerant (BFT) system that provides significant operational benefits: a plug-and-play system suitable for a broad set of hardware and network configurations, and adjusts effectively in real-time to changing fault scenarios and workloads. BFTBrain adapts to system conditions and application needs by switching between a set of BFT protocols in real-time. Two main advances contribute to BFTBrain's agility and performance. First, BFTBrain is based on a systematic, thorough modeling of metrics that correlate the performance of the studied BFT protocols with varying fault scenarios and workloads. These metrics are fed as features to BFTBrain's RL engine in order to choose the best-performing BFT protocols in real-time. Second, BFTBrain coordinates RL in a decentralized manner which is resilient to adversarial data pollution, where nodes share local metering values and reach the same learning output by consensus. As a result, in addition to providing significant operational benefits, BFTBrain improves throughput over fixed protocols by $18\%$ to $119\%$ under dynamic conditions and outperforms state-of-the-art learning based approaches by $44\%$ to $154\%$.

cs.DC↗

arXiv2024

Mitigate Negative Transfer with Similarity Heuristic Lifelong Prompt Tuning

Lifelong prompt tuning has significantly advanced parameter-efficient lifelong learning with its efficiency and minimal storage demands on various tasks. Our empirical studies, however, highlights certain transferability constraints in the current methodologies: a universal algorithm that guarantees consistent positive transfer across all tasks is currently unattainable, especially when dealing dissimilar tasks that may engender negative transfer. Identifying the misalignment between algorithm selection and task specificity as the primary cause of negative transfer, we present the Similarity Heuristic Lifelong Prompt Tuning (SHLPT) framework. This innovative strategy partitions tasks into two distinct subsets by harnessing a learnable similarity metric, thereby facilitating fruitful transfer from tasks regardless of their similarity or dissimilarity. Additionally, SHLPT incorporates a parameter pool to combat catastrophic forgetting effectively. Our experiments shows that SHLPT outperforms state-of-the-art techniques in lifelong learning benchmarks and demonstrates robustness against negative transfer in diverse task sequences.

cs.CL↗

arXiv2024

On Orchestrating Parallel Broadcasts for Distributed Ledgers

This paper introduces and develops the concept of ``ticketing'', through which atomic broadcasts are orchestrated by nodes in a distributed system. The paper studies different ticketing regimes that allow parallelism, yet prevent slow nodes from hampering overall progress. It introduces a hybrid scheme which combines managed and unmanaged ticketing regimes, striking a balance between adaptivity and resilience. The performance evaluation demonstrates how managed and unmanaged ticketing regimes benefit throughput in systems with heterogeneous resources both in static and dynamic scenarios, with the managed ticketing regime performing better among the two as it adapts better. Finally, it demonstrates how using the hybrid ticketing regime performance can enjoy both the adaptivity of the managed regime and the liveness guarantees of the unmanaged regime.

cs.DC↗

arXiv2023

Towards Learned Predictability of Storage Systems

With the rapid development of cloud computing and big data technologies, storage systems have become a fundamental building block of datacenters, incorporating hardware innovations such as flash solid state drives and non-volatile memories, as well as software infrastructures such as RAID and distributed file systems. Despite the growing popularity and interests in storage, designing and implementing reliable storage systems remains challenging, due to their performance instability and prevailing hardware failures. Proactive prediction greatly strengthens the reliability of storage systems. There are two dimensions of prediction: performance and failure. Ideally, through detecting in advance the slow IO requests, and predicting device failures before they really happen, we can build storage systems with especially low tail latency and high availability. While its importance is well recognized, such proactive prediction in storage systems, on the other hand, is particularly difficult. To move towards predictability of storage systems, various mechanisms and field studies have been proposed in the past few years. In this report, we present a survey of these mechanisms and field studies, focusing on machine learning based black-box approaches. Based on three representative research works, we discuss where and how machine learning should be applied in this field. The strengths and limitations of each research work are also evaluated in detail.

cs.DB↗

arXiv2023

AdaChain: A Learned Adaptive Blockchain

This paper presents AdaChain, a learning-based blockchain framework that adaptively chooses the best permissioned blockchain architecture in order to optimize effective throughput for dynamic transaction workloads. AdaChain addresses the challenge in the Blockchain-as-a-Service (BaaS) environments, where a large variety of possible smart contracts are deployed with different workload characteristics. AdaChain supports automatically adapting to an underlying, dynamically changing workload through the use of reinforcement learning. When a promising architecture is identified, AdaChain switches from the current architecture to the promising one at runtime in a way that respects correctness and security concerns. Experimentally, we show that AdaChain can converge quickly to optimal architectures under changing workloads, significantly outperform fixed architectures in terms of the number of successfully committed transactions, all while incurring low additional overhead.

cs.CR↗

arXiv2022

The Bedrock of Byzantine Fault Tolerance: A Unified Platform for BFT Protocol Design and Implementation

Byzantine Fault-Tolerant (BFT) protocols have recently been extensively used by decentralized data management systems with non-trustworthy infrastructures, e.g., permissioned blockchains. BFT protocols cover a broad spectrum of design dimensions from infrastructure settings such as the communication topology, to more technical features such as commitment strategy and even fundamental social choice properties like order-fairness. The proliferation of different BFT protocols has rendered it difficult to navigate the BFT landscape, let alone determine the protocol that best meets application needs. This paper presents Bedrock, a unified platform for BFT protocols design, analysis, implementation, and experiments. Bedrock proposes a design space consisting of a set of design choices capturing the trade-offs between different design space dimensions and providing fundamentally new insights into the strengths and weaknesses of BFT protocols. Bedrock enables users to analyze and experiment with BFT protocols within the space of plausible choices, evolve current protocols to design new ones, and even uncover previously unknown protocols. Our experimental results demonstrate the capability of Bedrock to uniformly evaluate BFT protocols in new ways that were not possible before due to the diverse assumptions made by these protocols. The results validate Bedrock's ability to analyze and derive BFT protocols.

cs.DC↗