AI alignment aims to make AI systems behave in line with human intentions and values. As
AI systems grow more capable, the potential large-scale risks associated with misaligned AI …

As a subfield of machine learning, reinforcement learning (RL) aims at optimizing decision
making by using interaction samples of an agent with its environment and the potentially …

Reinforcement learning from human feedback (RLHF) is a technique for training AI systems
to align with human goals. RLHF has emerged as the central method used to finetune state …

Artificial agents have traditionally been trained to maximize reward, which may incentivize
power-seeking and deception, analogous to how next-token prediction in language models …

Stable Diffusion revolutionised image creation from descriptive text. GPT-2, GPT-3 (. 5) and
GPT-4 demonstrated astonishing performance across a variety of language tasks. ChatGPT …

Reinforcement learning (RL) is a popular paradigm for addressing sequential decision tasks
in which the agent has only limited environmental feedback. Despite many advances over …

A prerequisite for social coordination is bidirectional communication between teammates,
each playing two roles simultaneously: as receptive listeners and expressive speakers. For …

Recent developments in the field of artificial intelligence (AI) have enabled new paradigms
of machine processing, shifting from data-driven, discriminative AI tasks toward …

Large language models (LLMs), designed to provide helpful and safe responses, often rely
on alignment techniques to align with user intent and social guidelines. Unfortunately, this …

Multiagent Reinforcement Learning (RL) solves complex tasks that require coordination with
other agents through autonomous exploration of the environment. However, learning a …