相关论文: State Contamination in Memory-Augmented LLM Agents

Persistent memory is turning language-model-based agents from stateless participants in isolated interactions into state-bearing components of LLM-based multi-agent systems. As memory becomes durable, reloadable, and behavior-shaping across…

Given the growing influence of language model-based agents on high-stakes societal decisions, from public policy to healthcare, ensuring their beneficial impact requires understanding the far-reaching implications of their suggestions. We…

Large language models (LLMs) are increasingly deployed as tool-using agents, shifting safety concerns from harmful text generation to harmful task completion. Deployed systems often condition on user profiles or persistent memory, yet agent…

As LLM-based agents are increasingly used in long-term interactions, cumulative memory is critical for enabling personalization and maintaining stylistic consistency. However, most existing systems adopt an ``all-or-nothing'' approach to…

Effectively processing long contexts remains a fundamental yet unsolved challenge for large language models (LLMs). Existing single-LLM-based methods primarily reduce the context window or optimize the attention mechanism, but they often…

As large language model (LLM)-powered agents are increasingly deployed to perform complex, real-world tasks, they face a growing class of attacks that exploit extended user-agent-environment interactions to pursue malicious objectives…

LLM agents deployed in production operate under operator-defined behavioral policies (system-prompt instructions such as prohibitions on credential disclosure, data exfiltration, and unauthorized output) that safety evaluations assume hold…

LLM agents have demonstrated remarkable performance across various applications, primarily due to their advanced capabilities in reasoning, utilizing external knowledge and tools, calling APIs, and executing actions to interact with…

Graph-based agent memory is increasingly used in LLM agents to support structured long-term recall and multi-hop reasoning, but it also creates a new poisoning surface: an attacker can inject a crafted relation into graph memory so that it…

Transformers suffer from a high computational cost that grows with sequence length for self-attention, making inference in long streams prohibited by memory consumption. Constant-memory alternatives such as RNNs and SSMs compress history…

Large language model (LLM) agents are fundamentally bottlenecked by finite context windows on long-horizon tasks. As trajectories grow, retaining tool outputs and intermediate reasoning in-context quickly becomes infeasible: the working…

LLM safety evaluations predominantly test models in isolation, yet deployed AI agents increasingly operate within persistent social environments alongside other agents. We introduce a Moltbook-style simulation platform where thousands of…

Memory makes LLM-based web agents personalized, powerful, yet exploitable. By storing past interactions to personalize future tasks, agents inadvertently create a persistent attack surface that spans websites and sessions. While existing…

Retrieval-Augmented Generation (RAG) systems have emerged as a promising solution to mitigate LLM hallucinations and enhance their performance in knowledge-intensive domains. However, these systems are vulnerable to adversarial poisoning…

Long-term memory is essential for persistent LLM agents, yet prevailing architectures store historical interactions as unstructured, flat text. This unconstrained storage induces provenance-role collapse, a critical failure mode where…

In reinforcement learning, an agent interacts sequentially with an environment to maximize a reward, receiving only partial, probabilistic feedback. This creates a fundamental exploration-exploitation trade-off: the agent must explore to…

The proliferation of open-weight Large Language Models (LLMs) has democratized agentic AI, yet fine-tuned weights are frequently shared and adopted with limited scrutiny beyond leaderboard performance. This creates a risk where third-party…

Modern language agents must operate over long-horizon, multi-turn interactions, where they retrieve external information, adapt to observations, and answer interdependent queries. Yet, most LLM systems rely on full-context prompting,…

Multi-agent AI pipelines typically assume that agent misconduct originates from model misalignment. We identify a structural failure in this assumption, the \emph{Misattribution Gap}, where memory-layer attacks produce behaviors…

Large Language Models (LLMs) have demonstrated remarkable capabilities across various tasks, yet they remain vulnerable to adversarial manipulations such as jailbreaking via prompt injection attacks. These attacks bypass safety mechanisms…