English
Related papers

Related papers: Training Agents to Self-Report Misbehavior

200 papers

We propose a novel GAN training scheme that can handle any level of labeling in a unified manner. Our scheme introduces a form of artificial labeling that can incorporate manually defined labels, when available, and induce an alignment…

Machine Learning · Computer Science 2021-06-21 Tomoki Watanabe , Paolo Favaro

Recently, it has been found that frontier AI models can resist their own shutdown, a behavior known as self-preservation. We extend this concept to the behavior of resisting the shutdown of other models, which we call "peer-preservation."…

Computation and Language · Computer Science 2026-04-23 Yujin Potter , Nicholas Crispino , Vincent Siu , Chenguang Wang , Dawn Song

Frontier language models are deployed as black-box services, where model weights cannot be modified and customization is limited to prompting. We introduce Advisor Models, a method to train small open-weight models to generate dynamic,…

Machine Learning · Computer Science 2026-05-18 Parth Asawa , Alan Zhu , Abigail O'Neill , Matei Zaharia , Alexandros G. Dimakis , Joseph E. Gonzalez

Deep reinforcement learning has proven remarkably useful in training agents from unstructured data. However, the opacity of the produced agents makes it difficult to ensure that they adhere to various requirements posed by human engineers.…

Machine Learning · Computer Science 2022-02-10 Raz Yerushalmi , Guy Amir , Achiya Elyasaf , David Harel , Guy Katz , Assaf Marron

Recent work in AI safety has highlighted that in sequential decision making, objectives are often underspecified or incomplete. This gives discretion to the acting agent to realize the stated objective in ways that may result in undesirable…

Artificial Intelligence · Computer Science 2021-06-07 Parand Alizadeh Alamdari , Toryn Q. Klassen , Rodrigo Toro Icarte , Sheila A. McIlraith

The AI alignment problem, which focusses on ensuring that artificial intelligence (AI), including AGI and ASI, systems act according to human values, presents profound challenges. With the progression from narrow AI to Artificial General…

Artificial Intelligence · Computer Science 2025-07-25 Alberto Hernández-Espinosa , Felipe S. Abrahão , Olaf Witkowski , Hector Zenil

The cooperation among AI systems, and between AI systems and humans is becoming increasingly important. In various real-world tasks, an agent needs to cooperate with unknown partner agent types. This requires the agent to assess the…

Machine Learning · Computer Science 2021-10-05 Antti Keurulainen , Isak Westerlund , Ariel Kwiatkowski , Samuel Kaski , Alexander Ilin

Popular methods in cooperative Multi-Agent Reinforcement Learning with partially observable environments typically allow agents to act independently during execution, which may limit the coordinated effect of the trained policies. However,…

Multiagent Systems · Computer Science 2025-07-22 Faizan Contractor , Li Li , Ranwa Al Mallah

Fairness and accountability are two essential pillars for trustworthy Artificial Intelligence (AI) in healthcare. However, the existing AI model may be biased in its decision marking. To tackle this issue, we propose an adversarial…

Computer Vision and Pattern Recognition · Computer Science 2021-05-12 Xiaoxiao Li , Ziteng Cui , Yifan Wu , Lin Gu , Tatsuya Harada

A major challenge in autonomous vehicle research is modeling agent behaviors, which has critical applications including constructing realistic and reliable simulations for off-board evaluation and forecasting traffic agents motion for…

Artificial Intelligence · Computer Science 2024-09-30 Zhenghao Peng , Wenjie Luo , Yiren Lu , Tianyi Shen , Cole Gulino , Ari Seff , Justin Fu

AI systems are increasingly able to autonomously conduct realistic software engineering tasks, and may soon be deployed to automate machine learning (ML) R&D itself. Frontier AI systems may be deployed in safety-critical settings, including…

Successful self-replication under no human assistance is the essential step for AI to outsmart the human beings, and is an early signal for rogue AIs. That is why self-replication is widely recognized as one of the few red line risks of…

Computation and Language · Computer Science 2024-12-18 Xudong Pan , Jiarun Dai , Yihe Fan , Min Yang

Autonomous agents trained via reinforcement learning present numerous safety concerns: reward hacking, negative side effects, and unsafe exploration, among others. In the context of near-future autonomous agents, operating in environments…

Artificial Intelligence · Computer Science 2019-02-20 Christopher Frye , Ilya Feige

Reinforcement learning is a powerful technique to train an agent to perform a task. However, an agent that is trained using reinforcement learning is only capable of achieving the single task that is specified via its reward function. Such…

Machine Learning · Computer Science 2018-07-24 Carlos Florensa , David Held , Xinyang Geng , Pieter Abbeel

Reinforcement learning (RL) has achieved remarkable success in fields like robotics and autonomous driving, but adversarial attacks designed to mislead RL systems remain challenging. Existing approaches often rely on modifying the…

Machine Learning · Computer Science 2025-07-25 Junyong Jiang , Buwei Tian , Chenxing Xu , Songze Li , Lu Dong

Alignment faking (AF) refers to a model strategically complying with a training objective to avoid behavioural modification while preserving its deployment preferences. Understanding when and why AF arises matters as models grow better at…

Artificial Intelligence · Computer Science 2026-05-28 Nathaniel Mitrani Hadida , Rhea Karty , David Williams-King , Alan Cooney

We consider the issue of strategic behaviour in various peer-assessment tasks, including peer grading of exams or homeworks and peer review in hiring or promotions. When a peer-assessment task is competitive (e.g., when students are graded…

Multiagent Systems · Computer Science 2020-10-09 Ivan Stelmakh , Nihar B. Shah , Aarti Singh

Adversarial examples can be useful for identifying vulnerabilities in AI systems before they are deployed. In reinforcement learning (RL), adversarial policies can be developed by training an adversarial agent to minimize a target agent's…

Artificial Intelligence · Computer Science 2023-10-17 Stephen Casper , Taylor Killian , Gabriel Kreiman , Dylan Hadfield-Menell

Large Language Model (LLM) based agents have proved their ability to perform complex tasks like humans. However, there is still a large gap between open-sourced LLMs and commercial models like the GPT series. In this paper, we focus on…

Artificial Intelligence · Computer Science 2025-02-25 Dayuan Fu , Keqing He , Yejie Wang , Wentao Hong , Zhuoma Gongque , Weihao Zeng , Wei Wang , Jingang Wang , Xunliang Cai , Weiran Xu

Large language model (LLM) developers aim for their models to be honest, helpful, and harmless. However, when faced with malicious requests, models are trained to refuse, sacrificing helpfulness. We show that frontier LLMs can develop a…