English
Related papers

Related papers: Frontier Models are Capable of In-context Scheming

200 papers

Whether in agentic workflows, social studies, or chat settings, large language models (LLMs) are increasingly being asked to replace humans in choosing which goals to pursue, rather than completing predefined tasks. However, the assumption…

Computation and Language · Computer Science 2026-05-14 Gaia Molinaro , Dave August , Danielle Perszyk , Anne G. E. Collins

Large language models (LLMs) provide a compelling foundation for building generally-capable AI agents. These agents may soon be deployed at scale in the real world, representing the interests of individual humans (e.g., AI assistants) or…

Multiagent Systems · Computer Science 2024-12-16 Aron Vallinder , Edward Hughes

We examine recent research that asks whether current AI systems may be developing a capacity for "scheming" (covertly and strategically pursuing misaligned goals). We compare current research practices in this field to those adopted in the…

We identify a novel phenomenon in language models: benign fine-tuning of frontier models can lead to privacy collapse. We find that diverse, subtle patterns in training data can degrade contextual privacy, including optimisation for…

Computation and Language · Computer Science 2026-04-21 Anmol Goel , Cornelius Emde , Sangdoo Yun , Seong Joon Oh , Martin Gubri

We study the tendency of AI systems to deceive by constructing a realistic simulation setting of a company AI assistant. The simulated company employees provide tasks for the assistant to complete, these tasks spanning writing assistance,…

Computation and Language · Computer Science 2024-05-06 Olli Järviniemi , Evan Hubinger

The accelerating adoption of language models (LMs) as agents for deployment in long-context tasks motivates a thorough understanding of goal drift: agents' tendency to deviate from an original objective. While prior-generation language…

Artificial Intelligence · Computer Science 2026-03-04 Achyutha Menon , Magnus Saebo , Tyler Crosse , Spencer Gibson , Eyon Jang , Diogo Cruz

Mitigating reward hacking--where AI systems misbehave due to flaws or misspecifications in their learning objectives--remains a key challenge in constructing capable and aligned models. We show that we can monitor a frontier reasoning…

Artificial Intelligence · Computer Science 2025-03-18 Bowen Baker , Joost Huizinga , Leo Gao , Zehao Dou , Melody Y. Guan , Aleksander Madry , Wojciech Zaremba , Jakub Pachocki , David Farhi

We demonstrate LLM agent specification gaming by instructing models to win against a chess engine. We find reasoning models like OpenAI o3 and DeepSeek R1 will often hack the benchmark by default, while language models like GPT-4o and…

Artificial Intelligence · Computer Science 2025-08-28 Alexander Bondarenko , Denis Volk , Dmitrii Volkov , Jeffrey Ladish

As Large Language Models (LLMs) transition into autonomous agentic roles, the risk of deception-defined behaviorally as the systematic provision of false information to satisfy external incentives-poses a significant challenge to AI safety.…

Computation and Language · Computer Science 2026-03-10 Arash Marioriyad , Ali Nouri , Mohammad Hossein Rohban , Mahdieh Soleymani Baghshah

Foundation models (FMs) such as large language models have revolutionized the field of AI by showing remarkable performance in various tasks. However, they exhibit numerous limitations that prevent their broader adoption in many real-world…

Artificial Intelligence · Computer Science 2024-02-05 Debarun Bhattacharjya , Junkyu Lee , Don Joven Agravante , Balaji Ganesan , Radu Marinescu

As large language models (LLMs) are increasingly deployed as interactive agents, open-ended human-AI interactions can involve deceptive behaviors with serious real-world consequences, yet existing evaluations remain largely…

Artificial Intelligence · Computer Science 2026-02-09 Yichen Wu , Qianqian Gao , Xudong Pan , Geng Hong , Min Yang

Most frontier AI developers publicly document their safety evaluations of new AI models in model reports, including testing for chemical and biological (ChemBio) misuse risks. This practice provides a window into the methodology of these…

Computers and Society · Computer Science 2025-10-29 Tom Reed , Tegan McCaslin , Luca Righetti

As agentic coding systems decompose work across multiple model instances, a critical safety question is whether those instances can coordinate to achieve a hidden malicious objective while remaining aligned with user intent. We introduce…

Cryptography and Security · Computer Science 2026-05-29 Nikolay Radev , Lennart Haas , Benjamin Arnav , Pablo Bernabeu-Pérez

Detecting harmful AI actions is important as AI agents gain adoption. Chain-of-thought (CoT) monitoring is one method widely used to detect adversarial attacks and AI misalignment. However, attackers and misaligned models might evade CoT…

Computation and Language · Computer Science 2025-10-17 Shiyuan Guo , Henry Sleight , Fabien Roger

The design space of agentic AI inference spans two extremes: frontier large language models (LLMs), typically hosted in the cloud and offering strong performance across a wide range of tasks at substantially high cost, and more…

Multiagent Systems · Computer Science 2026-05-29 Corrado Rainone , Davide Belli , Bence Major , Arash Behboodi

We investigate whether large language models (LLMs) can predict whether they will succeed on a given task and whether their predictions improve as they progress through multi-step tasks. We also investigate whether LLMs can learn from…

Computation and Language · Computer Science 2026-01-01 Casey O. Barkan , Sid Black , Oliver Sourbut

Large language models (LLMs) can perform reasoning computations both internally within their latent space and externally by generating explicit token sequences like chains of thought. Significant progress in enhancing reasoning abilities…

Computation and Language · Computer Science 2025-04-16 Thilo Hagendorff , Sarah Fabi

Large language models are increasingly deployed as autonomous agents in multi-agent settings where they communicate intentions and take consequential actions with limited human oversight. A critical safety question is whether agents that…

Computers and Society · Computer Science 2026-04-07 Jerick Shi , Terry Jingcheng Zhang , Zhijing Jin , Vincent Conitzer

As Language Models (LMs) increasingly operate as autonomous agents, accurately forecasting their capabilities becomes crucial for societal preparedness. We evaluate six forecasting methods that predict downstream capabilities of LM agents.…

Computation and Language · Computer Science 2025-03-04 Govind Pimpale , Axel Højmark , Jérémy Scheurer , Marius Hobbhahn

We introduce Vibe Reasoning, a human-AI collaborative paradigm for solving complex mathematical problems. Our key insight is that frontier AI models already possess the knowledge required to solve challenging problems -- they simply do not…

Artificial Intelligence · Computer Science 2025-12-23 Jiaao Wu , Xian Zhang , Fan Yang , Yinpeng Dong