243 citations · 443 across the 18 of their papers we have counts for
18 papers
Constitutional Classifiers++: Efficient Production-Grade Defenses against Universal Jailbreaks
Hoagy Cunningham, Jerry Wei, Zihan Wang +26
We introduce enhanced Constitutional Classifiers that deliver production-grade jailbreak robustness with dramatically reduced computational costs and refusal rates compared to prev…
Natural Emergent Misalignment from Reward Hacking in Production RL
Monte MacDiarmid, Benjamin Wright, Jonathan Uesato +19
We show that when large language models learn to reward hack on production RL environments, this can result in egregious emergent misalignment. We start with a pretrained model, im…
Human-AI Complementarity: A Goal for Amplified Oversight
Rishub Jain, Sophie Bridgers, Lili Janzer +3
Human feedback is critical for aligning AI systems to human values. As AI capabilities improve and AI is used to tackle more challenging tasks, verifying quality and safety becomes…
Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety
Tomek Korbak, Mikita Balesni, Elizabeth Barnes +38
AI systems that "think" in human language offer a unique opportunity for AI safety: we can monitor their chains of thought (CoT) for the intent to misbehave. Like all other known A…
Reasoning Models Don't Always Say What They Think
Yanda Chen, Joe Benton, Ansh Radhakrishnan +12
Chain-of-thought (CoT) offers a potential boon for AI safety as it allows monitoring a model's CoT to try to understand its intentions and reasoning processes. However, the effecti…
An Approach to Technical AGI Safety and Security
Rohin Shah, Alex Irpan, Alexander Matt Turner +27
Artificial General Intelligence (AGI) promises transformative benefits but also presents significant risks. We develop an approach to address the risk of harms consequential enough…
Challenges with unsupervised LLM knowledge discovery
Sebastian Farquhar, Vikrant Varma, Zachary Kenton +3
We show that existing unsupervised methods on large language model (LLM) activations do not discover knowledge -- instead they seem to discover whatever feature of the activations…
The Hydra Effect: Emergent Self-repair in Language Model Computations
Thomas McGrath, Matthew Rahtz, Janos Kramar +2
We investigate the internal structure of language model computations using causal analysis and demonstrate two motifs: (1) a form of adaptive computation where ablations of one att…
Does Circuit Analysis Interpretability Scale? Evidence from Multiple Choice Capabilities in Chinchilla
Tom Lieberum, Matthew Rahtz, János Kramár +4
\emph{Circuit analysis} is a promising technique for understanding the internal mechanisms of language models. However, existing analyses are done in small models far from the stat…
Tracr: Compiled Transformers as a Laboratory for Interpretability
David Lindner, János Kramár, Sebastian Farquhar +3
We show how to "compile" human-readable programs into standard decoder-only transformer models. Our compiler, Tracr, generates models with known structure. This structure can be us…
Teaching language models to support answers with verified quotes
Jacob Menick, Maja Trebacz, Vladimir Mikulik +8
Recent large language models often answer factual questions correctly. But users can't trust any given claim a model makes without fact-checking, because language models can halluc…
Scaling Language Models: Methods, Analysis & Insights from Training Gopher
Jack W. Rae, Sebastian Borgeaud, Trevor Cai +77
Language modelling provides a step towards intelligent communication systems by harnessing large repositories of written human knowledge to better predict and understand the world.…
Alignment of Language Agents
Zachary Kenton, Tom Everitt, Laura Weidinger +3
For artificial intelligence to be beneficial to humans the behaviour of AI agents needs to be aligned with what humans want. In this paper we discuss some behavioural issues for la…
Causal Analysis of Agent Behavior for AI Safety
Grégoire Déletang, Jordi Grau-Moya, Miljan Martic +6
As machine learning systems become more powerful they also become increasingly unpredictable and opaque. Yet, finding human-understandable explanations of how they work is essentia…
Algorithms for Causal Reasoning in Probability Trees
Tim Genewein, Tom McGrath, Grégoire Déletang +4
Probability trees are one of the simplest models of causal generative processes. They possess clean semantics and -- unlike causal Bayesian networks -- they can represent context-s…
Meta-trained agents implement Bayes-optimal agents
Vladimir Mikulik, Grégoire Delétang, Tom McGrath +4
Memory-based meta-learning is a powerful technique to build agents that adapt fast to any task within a target distribution. A previous theoretical study has argued that this remar…
Neural networks are a priori biased towards Boolean functions with low entropy
Chris Mingard, Joar Skalse, Guillermo Valle-Pérez +3
Understanding the inductive bias of neural networks is critical to explaining their ability to generalise. Here, for one of the simplest neural networks -- a single-layer perceptro…
Risks from Learned Optimization in Advanced Machine Learning Systems
Evan Hubinger, Chris van Merwijk, Vladimir Mikulik +2
We analyze the type of learned optimization that occurs when a learned model (such as a neural network) is itself an optimizer - a situation we refer to as mesa-optimization, a neo…