204 citations · 256 across the 26 of their papers we have counts for
32 papers
iS-KV: Online Low-Rank KV Cache Compression via Block-Incremental SVD
Yiren Zhao, Guanghui Song, Tianrui Qin +3
Long chain-of-thought reasoning substantially increases KV-cache memory during autoregressive decoding, as every generated token introduces new key and value states and causes the…
Grow the Harness, Not the Context: From Strategy-Free Scaffolds to Reusable Specialist Agents
Laizhen Li, Jiarui Li, Juanjuan Zhao +4
Large language model (LLM) agents often handle streams of related tasks, yet standard harnesses repeatedly ask the model to reconstruct the same control decisions inside each task'…
A2M: Trace-Optimized Agent Hijacking in the MCP Ecosystem
Laizhen Li, Xuan Wang, Peicheng Zhao +4
Agents using the Model Context Protocol (MCP) rely on semantic matching to select tools from third-party servers, exposing a semantic supply-chain risk through attacker-controlled…
CoCoScale: Leveraging Layer-wise Scaling to Unlock the Potential of Online LLM Serving
Jingfeng Wu, Yiyuan He, Minxian Xu +7
Online large language model (LLM) serving has become the backbone of modern AI applications, powering diverse downstream services through shared hardware clusters. However, modern…
Flash-Searcher: Fast and Effective Web Agents via DAG-Based Parallel Execution
Tianrui Qin, Qianben Chen, Sinuo Wang +10
Large language models (LLMs) have demonstrated remarkable capabilities in complex reasoning tasks when equipped with external tools. However, current frameworks predominantly rely…
BridgeNet: A Unified Multimodal Framework for Bridging 2D and 3D Industrial Anomaly Detection
An Xiang, Zixuan Huang, Xitong Gao +2
Industrial anomaly detection for 2D objects has gained significant attention and achieved progress in anomaly detection (AD) methods. However, identifying 3D depth anomalies using…
Unlock the Potential of Fine-grained LLM Serving via Dynamic Module Scaling
Jingfeng Wu, Yiyuan He, Minxian Xu +3
The rise of large language models (LLMs) has created new opportunities across various fields but has also introduced significant challenges in resource management. Current LLM serv…
OAgents: An Empirical Study of Building Effective Agents
He Zhu, Tianrui Qin, King Zhu +21
Recently, Agentic AI has become an increasingly popular research field. However, we argue that current agent research practices lack standardization and scientific rigor, making it…
Mixture of Weight-shared Heterogeneous Group Attention Experts for Dynamic Token-wise KV Optimization
Guanghui Song, Dongping Liao, Yiren Zhao +3
Transformer models face scalability challenges in causal language modeling (CLM) due to inefficient memory allocation for growing key-value (KV) caches, which strains compute and s…
Poison Once, Control Anywhere: Clean-Text Visual Backdoors in VLM-based Mobile Agents
Xuan Wang, Siyuan Liang, Zhe Liu +5
Mobile agents powered by vision-language models (VLMs) are increasingly adopted for tasks such as UI automation and camera-based assistance. These agents are typically fine-tuned u…
Offline Map Matching Based on Localization Error Distribution Modeling
Ruilin Xu, Yuchen Song, Kaijie Li +4
Offline map matching involves aligning historical trajectories of mobile objects, which may have positional errors, with digital maps. This is essential for applications in intelli…
Lie Detector: Unified Backdoor Detection via Cross-Examination Framework
Xuan Wang, Siyuan Liang, Dongping Liao +6
Institutions with limited data and computing resources often outsource model training to third-party providers in a semi-honest setting, assuming adherence to prescribed training p…
FLIP: Towards Comprehensive and Reliable Evaluation of Federated Prompt Learning
Dongping Liao, Xitong Gao, Yabo Xu +1
The increasing emphasis on privacy and data security has driven the adoption of federated learning, a decentralized approach to train machine learning models without sharing raw da…
Refining Salience-Aware Sparse Fine-Tuning Strategies for Language Models
Xinxin Liu, Aaron Thomas, Cheng Zhang +3
Parameter-Efficient Fine-Tuning (PEFT) has gained prominence through low-rank adaptation methods like LoRA. In this paper, we focus on sparsity-based PEFT (SPEFT), which introduces…
Scaling Laws For Mixed Quantization
Zeyu Cao, Boyang Gu, Cheng Zhang +5
Post-training quantization of Large Language Models (LLMs) has proven effective in reducing the memory and computational requirements for inference. In this study, we focus on a st…
Optimised Grouped-Query Attention Mechanism for Transformers
Yuang Chen, Cheng Zhang, Xitong Gao +3
Grouped-query attention (GQA) has been widely adopted in LLMs to mitigate the complexity of multi-head attention (MHA). To transform an MHA to a GQA, neighbour queries in MHA are e…
Unlocking the Global Synergies in Low-Rank Adapters
Zixi Zhang, Cheng Zhang, Xitong Gao +3
Low-rank Adaption (LoRA) has been the de-facto parameter-efficient fine-tuning technique for large language models. We present HeteroLoRA, a light-weight search algorithm that leve…
Fed-Credit: Robust Federated Learning with Credibility Management
Jiayan Chen, Zhirong Qian, Tianhui Meng +3
Aiming at privacy preservation, Federated Learning (FL) is an emerging machine learning approach enabling model training on decentralized devices or data sources. The learning mech…
APBench: A Unified Benchmark for Availability Poisoning Attacks and Defenses
Tianrui Qin, Xitong Gao, Juanjuan Zhao +2
The efficacy of availability poisoning, a method of poisoning data by injecting imperceptible perturbations to prevent its use in model training, has been a hot subject of investig…
Learning the Unlearnable: Adversarial Augmentations Suppress Unlearnable Example Attacks
Tianrui Qin, Xitong Gao, Juanjuan Zhao +2
Unlearnable example attacks are data poisoning techniques that can be used to safeguard public data against unauthorized use for training deep learning models. These methods add st…
Flareon: Stealthy any2any Backdoor Injection via Poisoned Augmentation
Tianrui Qin, Xianghuan He, Xitong Gao +3
Open software supply chain attacks, once successful, can exact heavy costs in mission-critical applications. As open-source ecosystems for deep learning flourish and become increas…
MORA: Improving Ensemble Robustness Evaluation with Model-Reweighing Attack
Yunrui Yu, Xitong Gao, Cheng-Zhong Xu
Adversarial attacks can deceive neural networks by adding tiny perturbations to their input data. Ensemble defenses, which are trained to minimize attack transferability among sub-…
Revisiting Structured Dropout
Yiren Zhao, Oluwatomisin Dada, Xitong Gao +1
Large neural networks are often overparameterised and prone to overfitting, Dropout is a widely used regularization technique to combat overfitting and improve model generalization…
Adversarial Attacks on ML Defense Models Competition
Yinpeng Dong, Qi-An Fu, Xiao Yang +25
Due to the vulnerability of deep neural networks (DNNs) to adversarial examples, a large number of defense techniques have been proposed to alleviate this problem in recent years.…