Related papers: LACUNA: Safe Agents as Recursive Program Holes
Large language model agents now act on codebases, browsers, operating systems, calendars, files, and tool ecosystems, but their evaluations often collapse behavior into final task success. AgentAtlas reframes agent evaluation as a…
Modern LLM agents solve complex tasks by operating in iterative execution loops, where they repeatedly reason, act, and self-evaluate progress to determine when a task is complete. In this work, we show that while this self-directed loop…
Self-healing systems have long been a focus of research, aiming to enable software to recover from unexpected runtime errors without human intervention. Traditional approaches rely on predefined heuristic rules, such as reusing error…
Large language models (LLMs) and their applications, such as agents, are highly vulnerable to prompt injection attacks. State-of-the-art prompt injection detection methods have the following limitations: (1) their effectiveness degrades…
Prompt injection attacks represent a major vulnerability in Large Language Model (LLM) deployments, where malicious instructions embedded in user inputs can override system prompts and induce unintended behaviors. This paper presents a…
Large language model (LLM) agents extend generative models with reasoning, tool use, and persistent memory, thereby enabling the automation of complex tasks. In healthcare, such systems could support documentation, care coordination, and…
Computer-use agents (CUAs) that interact with real computer systems can perform automated tasks but face critical safety risks. Ambiguous instructions may trigger harmful actions, and adversarial users can manipulate tool execution to…
Manufacturing industries are facing increasing product variability due to the growing demand for personalized products. Under these conditions, ensuring safety becomes challenging as frequent reconfigurations can lead to unintended…
Latent Action Models (LAMs) have emerged as an effective paradigm for handling heterogeneous datasets during Vision-Language-Action (VLA) model pretraining, offering a unified action space across embodiments. However, existing LAMs often…
Language model agents excel in long-session planning and reasoning, but existing benchmarks primarily focus on goal-oriented tasks with explicit objectives, neglecting creative adaptation in unfamiliar environments. To address this, we…
Agents for computer use (ACUs) are an emerging class of systems capable of executing complex tasks on digital devices -- such as desktops, mobile phones, and web platforms -- given instructions in natural language. These agents can automate…
In this paper, we present a challenging code reasoning task: vulnerability detection. Large Language Models (LLMs) have shown promising results in natural-language and math reasoning, but state-of-the-art (SOTA) models reported only 54.5%…
Deploying AI agents for repetitive periodic tasks exposes a critical tension: Large Language Models (LLMs) offer unmatched flexibility in tool orchestration, yet their inherent stochasticity causes unpredictable failures, and repeated…
Recent advances in AI agents capable of solving complex, everyday tasks, from scheduling to customer service, have enabled deployment in real-world settings, but their possibilities for unsafe behavior demands rigorous evaluation. While…
Reproducing computational research is often assumed to be as simple as rerunning the original code with provided data. In practice, missing packages, fragile file paths, version conflicts, or incomplete logic frequently cause analyses to…
High-privilege LLM agents that autonomously process external documentation are increasingly trusted to automate tasks by reading and executing project instructions, yet they are granted terminal access, filesystem control, and outbound…
The rapid proliferation of large language models (LLMs) has intensified the requirement for reliable safety evaluation to uncover model vulnerabilities. To this end, numerous LLM safety evaluation benchmarks are proposed. However, existing…
A conversation with a large language model (LLM) is a sequence of prompts and responses, with each response generated from the preceding conversation. AI agents build such conversations automatically: given an initial human prompt, a…
Agents aspire to eliminate the need for task-specific prompt crafting through autonomous reason-act-observe loops. Still, they are commonly instructed to follow a task-specific plan for guidance, e.g., to resolve software issues following…
This paper presents a secure-by-construction planning and control framework for multi-agent systems subject to linear temporal logic (LTL) specifications. The framework protects sensitive information from a passive intruder with partial…