15 papers
Resilience Matters for Embodied Agents System: New Metrics, Systematic Evaluation, and Optimization
Yapeng Liu, Yuanzhao Zhai, Xudong Gong +4
Embodied Agents System (EAS) are increasingly deployed in open-world physical domains, where reliability directly dictates deployment quality and human-agent trust. However, existi…
ReFrame: Evidence-Guided Test-Time Safety Alignment in Multimodal Large Language Models
Wenzheng Jiang, Xuankun Rong, Yuanzhao Zhai +2
While multimodal large language models (MLLMs) extend model capabilities beyond text, they also make safety alignment increasingly challenging. Multimodal safety alignment methods…
An Empirical Study on the Impact of Normalized Use-Case Specifications on Traceability
Luoyuan Shi, Yuanzhao Zhai, Dawei Feng +4
Traceability link recovery between requirements and source code is vital for software quality assurance and evolution analysis. Although automated traceability techniques have adva…
MCPEvol-Bench: Benchmarking LLM Agent Performance Across Dynamic Evolutions of MCP Servers
Huanxi Liu, Kun Hu, Jiaqi Liao +6
As Model Context Protocol (MCP) servers emerge as the core infrastructure for connecting LLMs with external tools, existing benchmarks leverage real-world MCP servers to evaluate L…
ParetoPilot: Zero-Surrogate Offline Multi-Objective Optimization via Infer-Perturb-Guide Diffusion
Ruiqing Sun, Sen Yang, Dawei Feng +3
Offline multi-objective optimization (Offline MOO) seeks Pareto-optimal designs from static datasets without additional environment interactions. Existing generative methods typica…
Beyond Scores: Diagnostic LLM Evaluation via Fine-Grained Abilities
Xu Zhang, Xudong Gong, Jiacheng Qin +5
Current evaluations of large language models aggregate performance across diverse tasks into single scores. This obscures fine-grained ability variation, limiting targeted model im…
Ranking Constraints via Topological Dual-Directional Search in Evolutionary Multi-Objective Optimization
Ruiqing Sun, Dawei Feng, Sheng Qi +6
Existing evolutionary algorithms for Constrained Multi-objective Optimization Problems (CMOPs) typically treat all constraints uniformly, overlooking their distinct geometric relat…
Decoupling Constraints from Two Directions for Evolutionary Constrained Multi-objective Optimization
Ruiqing Sun, Dawei Feng, Xing Zhou +6
Real-world constrained multi-objective optimization problems (CMOPs) commonly involve multiple constraints, and understanding and exploiting their coupling relationships is crucial…
Pay More Attention to the Robustness of Prompt for Instruction Data Mining
Qiang Wang, Dawei Feng, Xu Zhang +4
Instruction tuning has emerged as a paramount method for tailoring the behaviors of LLMs. Recent work has unveiled the potential for LLMs to achieve high performance through fine-t…
AudioCIL: A Python Toolbox for Audio Class-Incremental Learning with Multiple Scenes
Qisheng Xu, Yulin Sun, Yi Su +7
Deep learning, with its robust aotomatic feature extraction capabilities, has demonstrated significant success in audio signal processing. Typically, these methods rely on static,…
Exploring structure diversity in atomic resolution microscopy with graph neural networks
Zheng Luo, Ming Feng, Zijian Gao +9
The emergence of deep learning (DL) has provided great opportunities for the high-throughput analysis of atomic-resolution micrographs. However, the DL models trained by image patc…
AutoFeedback: An LLM-based Framework for Efficient and Accurate API Request Generation
Huanxi Liu, Jiaqi Liao, Dawei Feng +2
Large Language Models (LLMs) leverage external tools primarily through generating the API request to enhance task completion efficiency. The accuracy of API request generation sign…
Enhancing Decision-Making for LLM Agents via Step-Level Q-Value Models
Yuanzhao Zhai, Tingkai Yang, Kele Xu +4
Agents significantly enhance the capabilities of standalone Large Language Models (LLMs) by perceiving environments, making decisions, and executing actions. However, LLM agents st…
Online Self-Preferring Language Models
Yuanzhao Zhai, Zhuo Zhang, Kele Xu +6
Aligning with human preference datasets has been critical to the success of large language models (LLMs). Reinforcement learning from human feedback (RLHF) employs a costly reward…
Optimistic Model Rollouts for Pessimistic Offline Policy Optimization
Yuanzhao Zhai, Yiying Li, Zijian Gao +5
Model-based offline reinforcement learning (RL) has made remarkable progress, offering a promising avenue for improving generalization with synthetic model rollouts. Existing works…