Chen Tian

Zibo Wang, Yuhang Zhou, Zhibin Wang et al.

The increasing size of large language models (LLMs) has led to a surge in memory requirements during training, often exceeding the capacity of high-bandwidth memory (HBM). Swap-based memory optimization incurs neither accuracy loss nor additional end-to-end overhead when effectively overlapped, thus being an attractive solution. However, existing swap methods assume consistent operator sequences, which is impractical in Eager Mode, where operator sequences can vary during change. We propose Chameleon, which redesigns the end-to-end process of swap-based memory optimization and is the first work to consider varying operator sequences in Eager Mode. Chameleon (i) introduces a lightweight online profiler to enable continuous profiling for monitoring operator sequences, (ii) generates effective swap policies with limited operator information, and (iii) optimizes the policy execution module for accurate policy application and better performance. Experimental results demonstrate that Chameleon reduces profiling overhead by 84.25%, enables training models up to 4x larger than hardware memory while adapting to changes in operator sequences, improves performance by up to 38.94% compared to recomputation or high-degree parallelism.

Chenliang Tian, Zebo Yang, Raj Jain et al.

In quantum networks, one way to communicate is to distribute entanglements through swapping at intermediate nodes. Most existing work primarily aims to create efficient two-party end-to-end entanglement over long distances. However, some scenarios also require remote multipartite entanglement for applications such as quantum secret sharing and multi-party computation. Our previous study improved end-to-end entanglement rates using an asynchronous, tree-based routing scheme that relies solely on local knowledge of entanglement links, conserving unused entanglement and avoiding synchronous operations. This article extends this approach to multipartite entanglements, particularly the three-party Greenberger-Horne-Zeilinger (GHZ) states. It shows that our asynchronous protocol outperforms traditional synchronous methods in entanglement rates, especially as coherence times increase. This approach can also be extended to four-party and larger multipartite GHZ states, highlighting the effectiveness and adaptability of asynchronous routing for multipartite scenarios across various network topologies.

Yuhang Zhou, Zhibin Wang, Peng Jiang et al.

Training large language models faces frequent interruptions due to various faults, demanding robust fault-tolerance. Existing backup-free methods, such as redundant computation, dynamic parallelism, and data rerouting, each incur performance penalties, whether from ongoing overhead, lengthy reconfigurations, or post-recovery inefficiencies. We propose Chameleon, an adaptive fault-tolerant system that intelligently selects optimal recovery strategies when a failure occurs. Chameleon achieves this through a unified performance model, expedient execution plan search, accurate performance estimation, and efficient communication optimizations. Experiments on a 32-card cluster show that Chameleon maintains a performance gap of within 11.00% between post-recovery and failure-free training, while preserving model convergence and efficient memory usage. Compared to state-of-the-art methods, Chameleon achieves up to 1.229x and 1.355x higher average throughput than Oobleck and Recycle, respectively.

Chenliang Tian, Zebo Yang, Raj Jain et al.

Scalable quantum networks must support concurrent entanglement requests, yet existing routing protocols fail when users compete for shared repeater resources, wasting fragile quantum states. This paper presents RADAR-Q, a resource-aware decentralized routing protocol embedding real-time resource contention into path selection. Unlike prior designs requiring global coordination or central anchors, RADAR-Q makes intelligent local decisions balancing path length and fidelity, instantaneous quantum memory availability, and intermediate Bell-State Measurement (BSM) operations. By identifying the Nearest Common Ancestor (NCA) within a DODAG hierarchy, RADAR-Q localizes entanglement swapping close to communicating users - avoiding unnecessary central detours and reducing BSM chain length and decoherence exposure. We evaluate RADAR-Q on grid and random topologies against synchronous and root-centric asynchronous baselines. Results show RADAR-Q achieves aggregate throughputs 2.5x and 7.6x higher than synchronized and root-centric designs, respectively. While baselines suffer catastrophic fidelity collapse below the 0.5 threshold under high load, RADAR-Q consistently maintains end-to-end fidelity above 0.76, ensuring pairs remain usable. Furthermore, RADAR-Q exhibits near-perfect fairness (Jain's Fairness Index 96-98%) and retains over 50% of its ideal throughput under stringent 1.0 ms coherence times. These findings establish contention-aware decentralized routing as a scalable foundation for multi-tenant quantum networks.

Zhibin Wang, Shipeng Li, Xue Li et al.

Large language models have been widely deployed in various applications, encompassing both interactive online tasks and batched offline tasks. Given the burstiness and latency sensitivity of online tasks, over-provisioning resources is common practice. This allows for the integration of latency-insensitive offline tasks during periods of low online load, enhancing resource utilization. However, strategically serving online and offline tasks through a preemption mechanism fails to fully leverage the flexibility of offline tasks and suffers from KV cache recomputation and irregular workloads. In this paper, we introduce Echo, a collaborative online-offline task serving system, including a scheduler, a KV cache manager, and estimation toolkits. The scheduler and KV cache manager work tightly to maximize the throughput of offline tasks, while the estimator further predicts execution time to ensure online task SLOs. The scheduler leverages the batch information of last iteration to reduce the search space for finding the optimal schedule. The KV cache manager sets the priority of the KV cache based on the type of tasks and the opportunity of prefix sharing to reduce the recomputation. Finally, the estimation toolkits predict the execution time, future memory consumption, and the throughput of offline tasks to guide the scheduler, KV cache manager, and the system deployer. Evaluation based on real-world workloads demonstrates that Echo can increase offline task throughput by up to $3.3\times$, while satisfying online task SLOs.

Zhibin Wang, Zetao Hong, Xue Li et al.

Large Language Model (LLM) inference has emerged as a fundamental paradigm. In real-world scenarios, variations in output length cause severe workload imbalance in the decode phase, particularly for long-output reasoning tasks. Existing systems, such as PD disaggregation architectures, rely on static prefill-to-decode scheduling, which often results in SLO violations and OOM failures under evolving decode workloads. In this paper, we propose ARES, an adaptive decoding rescheduling system powered by length prediction to anticipate future workloads. Our core contributions include: (1) A lightweight and continuous LLM-native prediction method that leverages LLM hidden state to model remaining generation length with high precision (reducing MAE by 49.42%) and low overhead (cutting predictor parameters by 93.28%); (2) A rescheduling solution in decode phase with : A dynamic balancing mechanism that integrates current and predicted workloads, reducing P99 TPOT by 74.77% and achieving up to 2.24 times higher goodput.

Hongxu Pan, Shuxian Hu, Mo Zhou et al.

Researchers have proposed various methods of incorporating more structured information into the design of Graph Neural Networks (GNNs) to enhance their expressiveness. However, these methods are either computationally expensive or lacking in provable expressiveness. In this paper, we observe that the chords increase the complexity of the graph structure while contributing little useful information in many cases. In contrast, chordless structures are more efficient and effective for representing the graph. Therefore, when leveraging the information of cycles, we choose to omit the chords. Accordingly, we propose a Chordless Structure-based Graph Neural Network (CSGNN) and prove that its expressiveness is strictly more powerful than the k-hop GNN (KPGNN) with polynomial complexity. Experimental results on real-world datasets demonstrate that CSGNN outperforms existing GNNs across various graph tasks while incurring lower computational costs and achieving better performance than the GNNs of 3-WL expressiveness.

Liu Yafeng, Chen Tian, Liu Zhongyu et al.

Human actions recognition has attracted more and more people's attention. Many technology have been developed to express human action's features, such as image, skeleton-based, and channel state information(CSI). Among them, on account of CSI's easy to be equipped and undemanding for light, and it has gained more and more attention in some special scene. However, the relationship between CSI signal and human actions is very complex, and some preliminary work must be done to make CSI features easy to understand for computer. Nowadays, many work departed CSI-based features' action dealing into two parts. One part is for features extraction and dimension reduce, and the other part is for time series problems. Some of them even omitted one of the two part work. Therefore, the accuracies of current recognition systems are far from satisfactory. In this paper, we propose a new deep learning based approach, i.e. C3D network and C3D network with attention mechanism, for human actions recognition using CSI signals. This kind of network can make feature extraction from spatial convolution and temporal convolution simultaneously, and through this network the two part of CSI-based human actions recognition mentioned above can be realized at the same time. The entire algorithm structure is simplified. The experimental results show that our proposed C3D network is able to achieve the best recognition performance for all activities when compared with some benchmark approaches.

Juanjuan Zhao, Fan Zhang, Lai Tu et al.

Nowadays, metro systems play an important role in meeting the urban transportation demand in large cities. The understanding of passenger route choice is critical for public transit management. The wide deployment of Automated Fare Collection(AFC) systems opens up a new opportunity. However, only each trip's tap-in and tap-out timestamp and stations can be directly obtained from AFC system records; the train and route chosen by a passenger are unknown, which are necessary to solve our problem. While existing methods work well in some specific situations, they don't work for complicated situations. In this paper, we propose a solution that needs no additional equipment or human involvement than the AFC systems. We develop a probabilistic model that can estimate from empirical analysis how the passenger flows are dispatched to different routes and trains. We validate our approach using a large scale data set collected from the Shenzhen metro system. The measured results provide us with useful inputs when building the passenger path choice model.