Papers
Research papers from arXiv and related sources
Design-Specification Tiling for ICL-based CAD Code Generation
Large language models (LLMs) have demonstrated remarkable capabilities in code generation, yet they underperform on domain-specific tasks such as Computer-Aided Design (CAD) code generation due to ...
Yali Du, San-Zhuo Xi, Hui Sun, Ming Li
AI Planning Framework for LLM-Based Web Agents
Developing autonomous agents for web-based tasks is a core challenge in AI. While Large Language Model (LLM) agents can interpret complex user requests, they often operate as black boxes, making it...
Orit Shahnovsky, Rotem Dror
Cost-Efficient Multimodal LLM Inference via Cross-Tier GPU Heterogeneity
Multimodal large language model (MLLM) inference splits into two phases with opposing hardware demands: vision encoding is compute-bound, while language generation is memory-bandwidth-bound. We sho...
Donglin Yu
FGTR: Fine-Grained Multi-Table Retrieval via Hierarchical LLM Reasoning
With the rapid advancement of large language models (LLMs), growing efforts have been made on LLM-based table retrieval. However, existing studies typically focus on single-table query, and impleme...
Chaojie Sun, Bin Cao, Tiantian Li, Chenyu Hou, Ruizhe Li, Qing Fan
Seeing Eye to Eye: Enabling Cognitive Alignment Through Shared First-Person Perspective in Human-AI Collaboration
Despite advances in multimodal AI, current vision-based assistants often remain inefficient in collaborative tasks. We identify two key gulfs: a communication gulf, where users must translate rich ...
Zhuyu Teng, Pei Chen, Yichen Cai, Ruoqing Lu, Zhaoqu Jiang, Jiayang Li, Weitao You, Lingyun Sun
Experimental evidence of progressive ChatGPT models self-convergence
Large Language Models (LLMs) that undergo recursive training on synthetically generated data are susceptible to model collapse, a phenomenon marked by the generation of meaningless output. Existing...
Konstantinos F. Xylogiannopoulos, Petros Xanthopoulos, Panagiotis Karampelas, Georgios A. Bakamitsos
Colluding LoRA: A Composite Attack on LLM Safety Alignment
We introduce Colluding LoRA (CoLoRA), an attack in which each adapter appears benign and plausibly functional in isolation, yet their linear composition consistently compromises safety. Unlike atta...
Sihao Ding
MetaKE: Meta-learning Aligned Knowledge Editing via Bi-level Optimization
Knowledge editing (KE) aims to precisely rectify specific knowledge in Large Language Models (LLMs) without disrupting general capabilities. State-of-the-art methods suffer from an open-loop contro...
Shuxin Liu, Ou Wu
HyGra: Accelerating Network-State Simulation for LLM Training in DCNs via Adaptive Packet-Flow Granularity
In recent years, large language models (LLMs) have driven substantial intelligent transformation across diverse industries. Commercial LLM training is typically performed over data center networks ...
Wenyi Wang, Zheng Wu, Yanmeng Wang, Haolin Mao, Lei Han, Gaogang Xie, Fu Xiao
RetroReasoner: A Reasoning LLM for Strategic Retrosynthesis Prediction
Retrosynthesis prediction is a core task in organic synthesis that aims to predict reactants for a given product molecule. Traditionally, chemists select a plausible bond disconnection and derive c...
Hanbum Ko, Chanhui Lee, Ye Rin Kim, Rodrigo Hormazabal, Sehui Han, Sungbin Lim, Sungwoong Kim
From Text to Forecasts: Bridging Modality Gap with Temporal Evolution Semantic Space
Incorporating textual information into time-series forecasting holds promise for addressing event-driven non-stationarity; however, a fundamental modality gap hinders effective fusion: textual desc...
Lehui Li, Yuyao Wang, Jisheng Yan, Wei Zhang, Jinliang Deng, Haoliang Sun, Zhongyi Han, Yongshun ...
Evaluation of TCP Congestion Control for Public High-Performance Wide-Area Networks
Practitioners of a growing number of scientific and artificial-intelligence (AI) applications use High-Performance Wide-Area Networks (HP-WANs) for moving massive data sets between remote facilitie...
Fatih Berkay Sarpkaya, Andrea Francini, Bilgehan Erman, Shivendra Panwar
Continual Learning in Large Language Models: Methods, Challenges, and Opportunities
Continual learning (CL) has emerged as a pivotal paradigm to enable large language models (LLMs) to dynamically adapt to evolving knowledge and sequential tasks while mitigating catastrophic forget...
Hongyang Chen, Zhongwu Sun, Hongfei Ye, Kunchi Li, Xuemin Lin
A Standards-Aligned Coordination Framework for Edge-Enhanced Collaborative Healthcare in 6G Networks
Mission-critical healthcare applications including real-time intensive care monitoring, ambulance-to-hospital orchestration, and distributed medical imaging inference require workflow-level, time-b...
Liuwang Kang, Fan Wang, Yuzhang Huang, Shang Yan, Jianbin Zheng, Wenbin Lei, Konstantin Yakovlev,...
98$\times$ Faster LLM Routing Without a Dedicated GPU: Flash Attention, Prompt Compression, and Near-Streaming for the vLLM Semantic Router
System-level routers that intercept LLM requests for safety classification, domain routing, and PII detection must be both fast and operationally lightweight: they should add minimal latency to eve...
Xunzhuo Liu, Bowei He, Xue Liu, Andy Luo, Haichen Zhang, Huamin Chen
LightMoE: Reducing Mixture-of-Experts Redundancy through Expert Replacing
Mixture-of-Experts (MoE) based Large Language Models (LLMs) have demonstrated impressive performance and computational efficiency. However, their deployment is often constrained by substantial memo...
Jiawei Hao, Zhiwei Hao, Jianyuan Guo, Li Shen, Yong Luo, Han Hu, Dan Zeng
Uncovering Security Threats and Architecting Defenses in Autonomous Agents: A Case Study of OpenClaw
The rapid evolution of Large Language Models (LLMs) into autonomous, tool-calling agents has fundamentally altered the cybersecurity landscape. Frameworks like OpenClaw grant AI systems operating-s...
Zonghao Ying, Xiao Yang, Siyang Wu, Yumeng Song, Yang Qu, Hainan Li, Tianlin Li, Jiakai Wang, Ais...
RoboStereo: Dual-Tower 4D Embodied World Models for Unified Policy Optimization
Scalable Embodied AI faces fundamental constraints due to prohibitive costs and safety risks of real-world interaction. While Embodied World Models (EWMs) offer promise through imagined rollouts, e...
Ruicheng Zhang, Guangyu Chen, Zunnan Xu, Zihao Liu, Zhizhou Zhong, Mingyang Zhang, Jun Zhou, Xiu Li
Using a Human-AI Teaming Approach to Create and Curate Scientific Datasets with the SCILIRE System
The rapid growth of scientific literature has made manual extraction of structured knowledge increasingly impractical. To address this challenge, we introduce SCILIRE, a system for creating dataset...
Necva Bölücü, Jessica Irons, Changhyun Lee, Brian Jin, Maciej Rybinski, Huichen Yang, Andreas Due...
Spend Less, Reason Better: Budget-Aware Value Tree Search for LLM Agents
Test-time scaling has become a dominant paradigm for improving LLM agent reliability, yet current approaches treat compute as an abundant resource, allowing agents to exhaust token and tool budgets...
Yushu Li, Wenlong Deng, Jiajin Li, Xiaoxiao Li