7 papers
Lasso Games
Krzysztof Żyndul
We study games on finite graphs with labelled or weighted edges in which plays end as soon as the first loop is formed. The resulting play has a lasso shape: a path (head) followed…
Safety in Self-Evolving Agents: A Survey
Jiahao Chen, Zhou Feng, Oubo Ma +28
Large language models (LLMs) exhibit strong general capabilities, yet their parameters typically remain fixed after deployment, limiting learning from new interactions. In open-end…
WeSCE: A Benchmark for Measuring Security Drift in LLM-Driven Code Editing
Zhiyu Zhang, Tingyue Wen, Senke Sun +2
In this work, we introduce WeSCE, a benchmark for quantifying security drift in code editing under weak-security constraints, where tasks specify only functional objectives without…
VIPER-MCP: Detecting and Exploiting Taint-Style Vulnerabilities in Model Context Protocol Servers
Pengyu Sun, Zifeng Kang, Qishu Jin +4
Model Context Protocol (MCP) has emerged as a standard interface for connecting LLM agents to external tools. Because MCP servers expose privileged operations such as shell executi…
DMind-3: A Sovereign Edge--Local--Cloud AI System with Controlled Deliberation and Correction-Based Tuning for Safe, Low-Latency Transaction Execution
Enhao Huang, Frank Li, Tony Lin +1
This paper introduces DMind-3, a sovereign Edge-Local-Cloud intelligence stack designed to secure irreversible financial execution in Web3 environments against adversarial risks an…
Holographic Transformers for Complex-Valued Signal Processing: Integrating Phase Interference into Self-Attention
Enhao Huang, Zhiyu Zhang, Tianxiang Xu +6
Complex-valued signals encode both amplitude and phase, yet most deep models treat attention as real-valued correlation, overlooking interference effects. We introduce the Holograp…
DMind Benchmark: Toward a Holistic Assessment of LLM Capabilities across the Web3 Domain
Enhao Huang, Pengyu Sun, Shuxun Wang +13
The Web3 ecosystem, underpinned by cryptographic primitives and decentralized consensus, represents a high-stakes environment where software vulnerabilities and incentive misalignm…