12 papers
The Window Is in the Reasoning: When LLM Agents Decide to Follow Prompt Injections
Haoxuan Jia, Yang Liu, Bin Chong +13
LLM agents that read untrusted content can be hijacked by prompt injections. Runtime monitors that try to catch a hijack before the agent acts, and benchmarks that score each task…
Recasting the Destroy Step of Large Neighborhood Search as Dense Segmentation
Yang Liu, Yulin Huang, Jianshen Zhang +1
Large neighborhood search improves an incumbent by releasing selected variables and repairing the resulting subproblem under a time limit. FOVEA casts variable selection as dense s…
SabreAgent: Language Models at Design Time for Lost-Sales Inventory Control
Yang Liu, Yulin Huang, Xue Yu +3
SabreAgent uses a language model at design time to construct two components for lost-sales inventory control: a product-specific seasonal prior and a validation-selected capped bas…
Granularity-Adaptive Credit Assignment for Long-Horizon LLM Agent Reinforcement Learning
Taoran Liang, Yang Liu, Shang Luo +10
Long-horizon language-model agents trained with reinforcement learning oftenreceive sparse outcome rewards that do not reveal which decisions along a tra-jectory deserve credit. Ep…
Harness-agnostic detection and immunization of reward hacking in self-evolving language models
Rongxin Yang, Yang Liu, Shang Luo +10
Self-evolving language models improve by proposing candidate updates and keeping whatever raises a visible score. When that score is an imperfect proxy for the capability one actua…
Safety Does Not Compose: Non-Decaying Loop State for Autonomous LLM Agents
Chenhao Wu, Haoxuan Jia, Yang Liu +11
Large language model agents are increasingly deployed as autonomous loops. Starting from one human goal, such a system repeatedly discovers work, plans, executes tool calls, verifi…
Hindsight Memory-PRM: Supervising Memory Management with Auditable Hindsight Credit
Haoxuan Jia, Yang Liu, Yingguang Yang +14
Memory operations of long-horizon LLM agents are hard to supervise: an operation's value is unobservable when it is taken. But they are special -- they leave machine-readable evide…
Certified Multi-Turn Robustness for LLM Safety via Compositional Bounds and Safety Persistence
Yang Liu, Bin Chong, Wenkai Yang +9
Large language models (LLMs) are vulnerable to multi-turn jailbreak attacks that progressively manipulate conversation context. Existing certified robustness methods are limited to…
SafeHarness: Lifecycle-Integrated Security Architecture for LLM-based Agent Deployment
Xixun Lin, Yang Liu, Yancheng Chen +8
The performance of large language model (LLM) agents depends critically on the execution harness, the system layer that orchestrates tool use, context management, and state persist…
From Docs to Descriptions: Smell-Aware Evaluation of MCP Server Descriptions
Peiran Wang, Ying Li, Yuqiang Sun +3
The Model Context Protocol (MCP) has rapidly become a de facto standard for connecting LLM-based agents with external tools via reusable MCP servers. In practice, however, server s…
Descent-Guided Policy Gradient for Scalable Cooperative Multi-Agent Learning
Shan Yang, Yang Liu
Scaling cooperative multi-agent reinforcement learning (MARL) is fundamentally limited by cross-agent noise. When agents share a common reward, each agent's learning signal is comp…
Oyster-I: Beyond Refusal -- Constructive Safety Alignment for Responsible Language Models
Ranjie Duan, Jiexi Liu, Xiaojun Jia +27
Large language models (LLMs) typically deploy safety mechanisms to prevent harmful content generation. Most current approaches focus narrowly on risks posed by malicious actors, of…