相关论文: Guardrails Beat Guidance: A Large-Scale Study of R…
Software development, often perceived as a technical endeavor, is fundamentally a social activity requiring collaboration among team members. Acknowledging this, the software development community has devised strategies to address possible…
Large Language Model (LLM) agents are increasingly proposed to automate offensive security tasks, with recent studies reporting near human-level success rates in Capture-the-Flag (CTF) challenges. We here revisit these results, providing a…
Animals exhibit an innate ability to learn regularities of the world through interaction. By performing experiments in their environment, they are able to discern the causal factors of variation and infer how they affect the world's…
Context: In the workplace, some individuals engage in the voluntary and intentional generation, promotion, and realization of new ideas for the benefit of individual performance, group effectiveness, or the organization. The literature…
Preference-based reward learning is widely used for shaping agent behavior to match a user's preference, yet its sparse binary feedback makes it especially vulnerable to causal confusion. The learned reward often latches onto spurious…
Classical coding-theoretic guarantees often rely on trust assumptions, such as requiring sufficiently many honest nodes compared with adversarial ones. These assumptions are difficult to enforce in open decentralized systems where…
We study what actually works and what doesn't for training large language models as agents via multi-turn reinforcement learning. Despite rapid progress, existing frameworks and definitions are fragmented, and there is no systematic…
Software development is iterative, yet agentic coding benchmarks hide design issues through their single-shot setup. Recent iterative benchmarks attempt to remedy this but heavily constrain an agent's design decision space, making it…
AI coding scaffolds like Claude Code and Codex use retrying: blocking actions flagged as risky and continuing the trajectory. We study retrying from an AI control perspective, which treats the model as potentially adversarial. We find that…
Resolving team conflicts requires not only task-specific competence, but also social intelligence to find common ground and build consensus. As AI agents increasingly collaborate on complex work, they must develop coordination capabilities…
As machine learning models become more capable, they have exhibited increased potential in solving complex tasks. One of the most promising directions uses deep reinforcement learning to train autonomous agents in computer network defense…
As large language models (LLMs) evolve into autonomous agents that execute long-horizon workflows, invoking a high-capability model at every step becomes economically unsustainable. While model routing is effective for single-turn queries,…
Reinforcement learning (RL) algorithms, due to their reliance on external systems to learn from, require digital environments (e.g., simulators) with very simple interfaces, which in turn constrain significantly the implementation of such…
Quadratic programs arise in robotics, communications, smart grids, and many other applications. As these problems grow in size, finding solutions becomes more computationally demanding, and new algorithms are needed to efficiently solve…
Coding agents are rapidly changing the landscape of software development, moving from inline completion to autonomous systems that edit repositories, open pull requests, respond to issues, and run scheduled or webhook triggered routines…
In the first half of 2025, coding agents have emerged as a category of development tools that have very quickly transitioned to the practice. Unlike ''traditional'' code completion LLMs such as Copilot, agents like Cursor, Claude Code, or…
Reinforcement Learning is a highly active research field with promising advancements. In the field of autonomous driving, however, often very simple scenarios are being examined. Common approaches use non-interpretable control commands as…
Insecure default values in software settings can be exploited by attackers to compromise the system that runs the software. As a countermeasure, there exist security-configuration guides specifying in detail which values are secure.…
Large Language Model (LLM) Agents are advancing quickly, with the increasing leveraging of LLM Agents to assist in development tasks such as code generation. While LLM Agents accelerate code generation, studies indicate they may introduce…
This article presents GuideR, a user-guided rule induction algorithm, which overcomes the largest limitation of the existing methods-the lack of the possibility to introduce user's preferences or domain knowledge to the rule learning…