Related papers: Do Coding Agents Understand Least-Privilege Author…
HyperLTL, the extension of Linear Temporal Logic by trace quantifiers, is a uniform framework for expressing information flow policies by relating multiple traces of a security-critical system. HyperLTL has been successfully applied to…
Model attribution is a critical component of deep neural networks (DNNs) for its interpretability to complex models. Recent studies bring up attention to the security of attribution methods as they are vulnerable to attribution attacks that…
Observers and practitioners of artificial intelligence (AI) have proposed an FDA-style licensing regime for the most advanced AI models, or 'frontier' models. In this paper, we explore the applicability of approval regulation -- that is,…
Many interventions, such as vaccines in clinical trials or coupons in online marketplaces, must be assigned sequentially without full knowledge of their effects. Multi-armed bandit algorithms have proven successful in such settings.…
Unreadable code could be a breeding ground for errors. Thus, previous work defined approaches based on machine learning to automatically assess code readability that can warn developers when some code artifacts (e.g., classes) become…
Current LLM agent benchmarks, which predominantly focus on binary pass/fail tasks such as code generation or search-based question answering, often neglect the value of real-world engineering that is often captured through the iterative…
Claude Code's auto mode is the first deployed permission system for AI coding agents, using a two-stage transcript classifier to gate dangerous tool calls. Anthropic reports a 0.4% false positive rate and 17% false negative rate on…
Agent evaluation requires assessing complex multi-step behaviors involving tool use and intermediate reasoning, making it costly and expertise-intensive. A natural question arises: can frontier coding assistants reliably automate this…
User authorization queries in the context of role-based access control have attracted considerable interest in the last 15 years. Such queries are used to determine whether it is possible to allocate a set of roles to a user that enables…
LLM agents increasingly run inside execution harnesses that dispatch tools, allocate resources, and route messages between specialized components. However, a harness can return a correct, benign answer over a trajectory that accesses…
Language models as intelligent agents push the boundaries of sequential decision-making agents but struggle with limited knowledge of environmental dynamics and exponentially huge action space. Recent efforts like GLAM and TWOSOME manually…
Training large language model agents on tasks at the frontier of their capabilities is key to unlocking advanced reasoning. We introduce a data synthesis approach inspired by the educational theory of the Zone of Proximal Development (ZPD),…
LLM-based agents are increasingly deployed in high-stakes scenarios such as email management, financial transactions, and code execution, where they interact with the external world through tool calling. During execution, these agents must…
Tool calling agents are an emerging paradigm in LLM deployment, with major platforms such as ChatGPT, Claude, and Gemini adding connectors and autonomous capabilities. However, the inherent unreliability of LLMs introduces fundamental…
In recent years, the AI wave has grown rapidly in software development. Even novice developers can now design and generate complex framework-constrained software systems based on their high-level requirements with the help of Large Language…
In reinforcement learning, we typically refer to unsupervised pre-training when we aim to pre-train a policy without a priori access to the task specification, i.e. rewards, to be later employed for efficient learning of downstream tasks.…
Existing Agent benchmarks suffer from two critical limitations: high environment interaction overhead (up to 41\% of total evaluation time) and imbalanced task horizon and difficulty distributions that make aggregate scores unreliable. To…
As autonomous coding agents become deeply embedded in software development workflows, their high operational velocity introduces a critical oversight challenge: the accumulating divergence between agentic actions and architectural intent.…
Regulators currently govern the AI data economy based on intuition rather than evidence, struggling to choose between inconsistent regimes of informed consent, immunity, and liability. To fill this policy vacuum, this paper develops a novel…
Deep neural networks do not discriminate between spurious and causal patterns, and will only learn the most predictive ones while ignoring the others. This shortcut learning behaviour is detrimental to a network's ability to generalize to…