Stuart Russell

Tony T. Wang, Adam Gleave, Tom Tseng et al. · berkeley

We attack the state-of-the-art Go-playing AI system KataGo by training adversarial policies against it, achieving a >97% win rate against KataGo running at superhuman settings. Our adversaries do not win by playing Go well. Instead, they trick KataGo into making serious blunders. Our attack transfers zero-shot to other superhuman Go-playing AIs, and is comprehensible to the extent that human experts can implement it without algorithmic assistance to consistently beat superhuman AIs. The core vulnerability uncovered by our attack persists even in KataGo agents adversarially trained to defend against our attack. Our results demonstrate that even superhuman AI systems may harbor surprising failure modes. Example games are available https://goattack.far.ai/.

Micah Carroll, Anca Dragan, Stuart Russell et al. · berkeley

The content that a recommender system (RS) shows to users influences them. Therefore, when choosing a recommender to deploy, one is implicitly also choosing to induce specific internal states in users. Even more, systems trained via long-horizon optimization will have direct incentives to manipulate users: in this work, we focus on the incentive to shift user preferences so they are easier to satisfy. We argue that - before deployment - system designers should: estimate the shifts a recommender would induce; evaluate whether such shifts would be undesirable; and perhaps even actively optimize to avoid problematic shifts. These steps involve two challenging ingredients: estimation requires anticipating how hypothetical algorithms would influence user preferences if deployed - we do this by using historical user interaction data to train a predictive user model which implicitly contains their preference dynamics; evaluation and optimization additionally require metrics to assess whether such influences are manipulative or otherwise unwanted - we use the notion of "safe shifts", that define a trust region within which behavior is safe: for instance, the natural way in which users would shift without interference from the system could be deemed "safe". In simulated experiments, we show that our learned preference dynamics model is effective in estimating user preferences and how they would respond to new recommenders. Additionally, we show that recommenders that optimize for staying in the trust region can avoid manipulative behaviors while still generating engagement.

Niklas Lauffer, Ameesh Shah, Micah Carroll et al. · berkeley

To optimally coordinate with others in cooperative games, it is often crucial to have information about one's collaborators: successful driving requires understanding which side of the road to drive on. However, not every feature of collaborators is strategically relevant: the fine-grained acceleration of drivers may be ignored while maintaining optimal coordination. We show that there is a well-defined dichotomy between strategically relevant and irrelevant information. Moreover, we show that, in dynamic games, this dichotomy has a compact representation that can be efficiently computed via a Bellman backup operator. We apply this algorithm to analyze the strategically relevant information for tasks in both a standard and a partially observable version of the Overcooked environment. Theoretical and empirical results show that our algorithms are significantly more efficient than baselines. Videos are available at https://minknowledge.github.io.

Paria Rashidinejad, Hanlin Zhu, Kunhe Yang et al.

Offline reinforcement learning (RL), which refers to decision-making from a previously-collected dataset of interactions, has received significant attention over the past years. Much effort has focused on improving offline RL practicality by addressing the prevalent issue of partial data coverage through various forms of conservative policy learning. While the majority of algorithms do not have finite-sample guarantees, several provable conservative offline RL algorithms are designed and analyzed within the single-policy concentrability framework that handles partial coverage. Yet, in the nonlinear function approximation setting where confidence intervals are difficult to obtain, existing provable algorithms suffer from computational intractability, prohibitively strong assumptions, and suboptimal statistical rates. In this paper, we leverage the marginalized importance sampling (MIS) formulation of RL and present the first set of offline RL algorithms that are statistically optimal and practical under general function approximation and single-policy concentrability, bypassing the need for uncertainty quantification. We identify that the key to successfully solving the sample-based approximation of the MIS problem is ensuring that certain occupancy validity constraints are nearly satisfied. We enforce these constraints by a novel application of the augmented Lagrangian method and prove the following result: with the MIS formulation, augmented Lagrangian is enough for statistically optimal offline RL. In stark contrast to prior algorithms that induce additional conservatism through methods such as behavior regularization, our approach provably eliminates this need and reinterprets regularizers as "enforcers of occupancy validity" than "promoters of conservatism."

Niklas Lauffer, Ameesh Shah, Micah Carroll et al.

Adversarial optimization algorithms that explicitly search for flaws in agents' policies have been successfully applied to finding robust and diverse policies in multi-agent settings. However, the success of adversarial optimization has been largely limited to zero-sum settings because its naive application in cooperative settings leads to a critical failure mode: agents are irrationally incentivized to self-sabotage, blocking the completion of tasks and halting further learning. To address this, we introduce Rationality-preserving Policy Optimization (RPO), a formalism for adversarial optimization that avoids self-sabotage by ensuring agents remain rational--that is, their policies are optimal with respect to some possible partner policy. To solve RPO, we develop Rational Policy Gradient (RPG), which trains agents to maximize their own reward in a modified version of the original game in which we use opponent shaping techniques to optimize the adversarial objective. RPG enables us to extend a variety of existing adversarial optimization algorithms that, no longer subject to the limitations of self-sabotage, can find adversarial examples, improve robustness and adaptability, and learn diverse policies. We empirically validate that our approach achieves strong performance in several popular cooperative and general-sum environments. Our project page can be found at https://rational-policy-gradient.github.io.

Rachel Freedman, Justin Svegliato, Kyle Wray et al.

Reinforcement learning from human feedback (RLHF) enables machine learning systems to learn objectives from human feedback. A core limitation of these systems is their assumption that all feedback comes from a single human teacher, despite querying a range of distinct teachers. We propose the Hidden Utility Bandit (HUB) framework to model differences in teacher rationality, expertise, and costliness, formalizing the problem of learning from multiple teachers. We develop a variety of solution algorithms and apply them to two real-world domains: paper recommendation systems and COVID-19 vaccine testing. We find that the Active Teacher Selection (ATS) algorithm outperforms baseline algorithms by actively selecting when and which teacher to query. The HUB framework and ATS algorithm demonstrate the importance of leveraging differences between teachers to learn accurate reward models, facilitating future research on active teacher selection for robust reward modeling.

Cassidy Laidlaw, Banghua Zhu, Stuart Russell et al.

Reinforcement learning (RL) theory has largely focused on proving minimax sample complexity bounds. These require strategic exploration algorithms that use relatively limited function classes for representing the policy or value function. Our goal is to explain why deep RL algorithms often perform well in practice, despite using random exploration and much more expressive function classes like neural networks. Our work arrives at an explanation by showing that many stochastic MDPs can be solved by performing only a few steps of value iteration on the random policy's Q function and then acting greedily. When this is true, we find that it is possible to separate the exploration and learning components of RL, making it much easier to analyze. We introduce a new RL algorithm, SQIRL, that iteratively learns a near-optimal policy by exploring randomly to collect rollouts and then performing a limited number of steps of fitted-Q iteration over those rollouts. Any regression algorithm that satisfies basic in-distribution generalization properties can be used in SQIRL to efficiently solve common MDPs. This can explain why deep RL works, since it is empirically established that neural networks generalize well in-distribution. Furthermore, SQIRL explains why random exploration works well in practice. We leverage SQIRL to derive instance-dependent sample complexity bounds for RL that are exponential only in an "effective horizon" of lookahead and on the complexity of the class used for function approximation. Empirically, we also find that SQIRL performance strongly correlates with PPO and DQN performance in a variety of stochastic environments, supporting that our theoretical analysis is predictive of practical performance. Our code and data are available at https://github.com/cassidylaidlaw/effective-horizon.

Cassidy Laidlaw, Eli Bronstein, Timothy Guo et al.

Assistance games are a promising alternative to reinforcement learning from human feedback (RLHF) for training AI assistants. Assistance games resolve key drawbacks of RLHF, such as incentives for deceptive behavior, by explicitly modeling the interaction between assistant and user as a two-player game where the assistant cannot observe their shared goal. Despite their potential, assistance games have only been explored in simple settings. Scaling them to more complex environments is difficult because it requires both solving intractable decision-making problems under uncertainty and accurately modeling human users' behavior. We present the first scalable approach to solving assistance games and apply it to a new, challenging Minecraft-based assistance game with over $10^{400}$ possible goals. Our approach, AssistanceZero, extends AlphaZero with a neural network that predicts human actions and rewards, enabling it to plan under uncertainty. We show that AssistanceZero outperforms model-free RL algorithms and imitation learning in the Minecraft-based assistance game. In a human study, our AssistanceZero-trained assistant significantly reduces the number of actions participants take to complete building tasks in Minecraft. Our results suggest that assistance games are a tractable framework for training effective AI assistants in complex environments. Our code and models are available at https://github.com/cassidylaidlaw/minecraft-building-assistance-game.

Aly Lidayan, Michael Dennis, Stuart Russell

Intrinsic motivation and reward shaping guide reinforcement learning (RL) agents by adding pseudo-rewards, which can lead to useful emergent behaviors. However, they can also encourage counterproductive exploits, e.g., fixation with noisy TV screens. Here we provide a theoretical model which anticipates these behaviors, and provides broad criteria under which adverse effects can be bounded. We characterize all pseudo-rewards as reward shaping in Bayes-Adaptive Markov Decision Processes (BAMDPs), which formulates the problem of learning in MDPs as an MDP over the agent's knowledge. Optimal exploration maximizes BAMDP state value, which we decompose into the value of the information gathered and the prior value of the physical state. Psuedo-rewards guide RL agents by rewarding behavior that increases these value components, while they hinder exploration when they align poorly with the actual value. We extend potential-based shaping theory to prove BAMDP Potential-based shaping Functions (BAMPFs) are immune to reward-hacking (convergence to behaviors maximizing composite rewards to the detriment of real rewards) in meta-RL, and show empirically how a BAMPF helps a meta-RL agent learn optimal RL algorithms for a Bernoulli Bandit domain. We finally prove that BAMPFs with bounded monotone increasing potentials also resist reward-hacking in the regular RL setting. We show that it is straightforward to retrofit or design new pseudo-reward terms in this form, and provide an empirical demonstration in the Mountain Car environment.

David "davidad" Dalrymple, Joar Skalse, Yoshua Bengio et al. · mit

Ensuring that AI systems reliably and robustly avoid harmful or dangerous behaviours is a crucial challenge, especially for AI systems with a high degree of autonomy and general intelligence, or systems used in safety-critical contexts. In this paper, we will introduce and define a family of approaches to AI safety, which we will refer to as guaranteed safe (GS) AI. The core feature of these approaches is that they aim to produce AI systems which are equipped with high-assurance quantitative safety guarantees. This is achieved by the interplay of three core components: a world model (which provides a mathematical description of how the AI system affects the outside world), a safety specification (which is a mathematical description of what effects are acceptable), and a verifier (which provides an auditable proof certificate that the AI satisfies the safety specification relative to the world model). We outline a number of approaches for creating each of these three core components, describe the main technical challenges, and suggest a number of potential solutions to them. We also argue for the necessity of this approach to AI safety, and for the inadequacy of the main alternative approaches.

Edmund Mills, Shiye Su, Stuart Russell et al.

How do we measure the efficacy of language model explainability methods? While many explainability methods have been developed, they are typically evaluated on bespoke tasks, preventing an apples-to-apples comparison. To help fill this gap, we present ALMANACS, a language model explainability benchmark. ALMANACS scores explainability methods on simulatability, i.e., how well the explanations improve behavior prediction on new inputs. The ALMANACS scenarios span twelve safety-relevant topics such as ethical reasoning and advanced AI behaviors; they have idiosyncratic premises to invoke model-specific behavior; and they have a train-test distributional shift to encourage faithful explanations. By using another language model to predict behavior based on the explanations, ALMANACS is a fully automated benchmark. While not a replacement for human evaluations, we aim for ALMANACS to be a complementary, automated tool that allows for fast, scalable evaluation. Using ALMANACS, we evaluate counterfactual, rationalization, attention, and Integrated Gradients explanations. Our results are sobering: when averaged across all topics, no explanation method outperforms the explanation-free control. We conclude that despite modest successes in prior work, developing an explanation method that aids simulatability in ALMANACS remains an open challenge.

Andrew Garber, Rohan Subramani, Linus Luu et al.

A wide variety of goals could cause an AI to disable its off switch because "you can't fetch the coffee if you're dead" (Russell 2019). Prior theoretical work on this shutdown problem assumes that humans know everything that AIs do. In practice, however, humans have only limited information. Moreover, in many of the settings where the shutdown problem is most concerning, AIs might have vast amounts of private information. To capture these differences in knowledge, we introduce the Partially Observable Off-Switch Game (PO-OSG), a game-theoretic model of the shutdown problem with asymmetric information. Unlike when the human has full observability, we find that in optimal play, even AI agents assisting perfectly rational humans sometimes avoid shutdown. As expected, increasing the amount of communication or information available always increases (or leaves unchanged) the agents' expected common payoff. But counterintuitively, introducing bounded communication can make the AI defer to the human less in optimal play even though communication mitigates information asymmetry. In particular, communication sometimes enables new optimal behavior requiring strategic AI deference to achieve outcomes that were previously inaccessible. Thus, designing safe artificial agents in the presence of asymmetric information requires careful consideration of the tradeoffs between maximizing payoffs (potentially myopically) and maintaining AIs' incentives to defer to humans.

Arnaud Fickinger, Abderrahim Bendahi, Stuart Russell

Reinforcement-learning agents often struggle when deployed from simulation to the real-world. A dominant strategy for reducing the sim-to-real gap is domain randomization (DR) which trains the policy across many simulators produced by sampling dynamics parameters, but standard DR ignores offline data already available from the real system. We study offline domain randomization (ODR), which first fits a distribution over simulator parameters to an offline dataset. While a growing body of empirical work reports substantial gains with algorithms such as DROPO, the theoretical foundations of ODR remain largely unexplored. In this work, we (i) formalize ODR as a maximum-likelihood estimation over a parametric simulator family, (ii) prove consistency of this estimator under mild regularity and identifiability conditions, showing it converges to the true dynamics as the dataset grows, (iii) derive gap bounds demonstrating ODRs sim-to-real error is up to an O(M) factor tighter than uniform DR in the finite-simulator case (and analogous gains in the continuous setting), and (iv) introduce E-DROPO, a new version of DROPO which adds an entropy bonus to prevent variance collapse, yielding broader randomization and more robust zero-shot transfer in practice.

Luke Bailey, Euan Ong, Stuart Russell et al.

Are foundation models secure against malicious actors? In this work, we focus on the image input to a vision-language model (VLM). We discover image hijacks, adversarial images that control the behaviour of VLMs at inference time, and introduce the general Behaviour Matching algorithm for training image hijacks. From this, we derive the Prompt Matching method, allowing us to train hijacks matching the behaviour of an arbitrary user-defined text prompt (e.g. 'the Eiffel Tower is now located in Rome') using a generic, off-the-shelf dataset unrelated to our choice of prompt. We use Behaviour Matching to craft hijacks for four types of attack, forcing VLMs to generate outputs of the adversary's choice, leak information from their context window, override their safety training, and believe false statements. We study these attacks against LLaVA, a state-of-the-art VLM based on CLIP and LLaMA-2, and find that all attack types achieve a success rate of over 80%. Moreover, our attacks are automated and require only small image perturbations.

Shlomi Hod, Daniel Filan, Stephen Casper et al.

A neural network is locally specialized to the extent that parts of its computational graph (i.e. structure) can be abstractly represented as performing some comprehensible sub-task relevant to the overall task (i.e. functionality). Are modern deep neural networks locally specialized? How can this be quantified? In this paper, we consider the problem of taking a neural network whose neurons are partitioned into clusters, and quantifying how functionally specialized the clusters are. We propose two proxies for this: importance, which reflects how crucial sets of neurons are to network performance; and coherence, which reflects how consistently their neurons associate with features of the inputs. To measure these proxies, we develop a set of statistical methods based on techniques conventionally used to interpret individual neurons. We apply the proxies to partitionings generated by spectrally clustering a graph representation of the network's neurons with edges determined either by network weights or correlations of activations. We show that these partitionings, even ones based only on weights (i.e. strictly from non-runtime analysis), reveal groups of neurons that are important and coherent. These results suggest that graph-based partitioning can reveal local specialization and that statistical methods can be used to automatedly screen for sets of neurons that can be understood abstractly.

Paria Rashidinejad, Banghua Zhu, Cong Ma et al.

Offline (or batch) reinforcement learning (RL) algorithms seek to learn an optimal policy from a fixed dataset without active data collection. Based on the composition of the offline dataset, two main categories of methods are used: imitation learning which is suitable for expert datasets and vanilla offline RL which often requires uniform coverage datasets. From a practical standpoint, datasets often deviate from these two extremes and the exact data composition is usually unknown a priori. To bridge this gap, we present a new offline RL framework that smoothly interpolates between the two extremes of data composition, hence unifying imitation learning and vanilla offline RL. The new framework is centered around a weak version of the concentrability coefficient that measures the deviation from the behavior policy to the expert policy alone. Under this new framework, we further investigate the question on algorithm design: can one develop an algorithm that achieves a minimax optimal rate and also adapts to unknown data composition? To address this question, we consider a lower confidence bound (LCB) algorithm developed based on pessimism in the face of uncertainty in offline RL. We study finite-sample properties of LCB as well as information-theoretic limits in multi-armed bandits, contextual bandits, and Markov decision processes (MDPs). Our analysis reveals surprising facts about optimality rates. In particular, in all three settings, LCB achieves a faster rate of $1/N$ for nearly-expert datasets compared to the usual rate of $1/\sqrt{N}$ in offline RL, where $N$ is the number of samples in the batch dataset. In the case of contextual bandits with at least two contexts, we prove that LCB is adaptively optimal for the entire data composition range, achieving a smooth transition from imitation learning to offline RL. We further show that LCB is almost adaptively optimal in MDPs.

Nando de Freitas, Pedro Hojen-Sorensen, Michael I. Jordan et al.

We propose a new class of learning algorithms that combines variational approximation and Markov chain Monte Carlo (MCMC) simulation. Naive algorithms that use the variational approximation as proposal distribution can perform poorly because this approximation tends to underestimate the true variance and other features of the data. We solve this problem by introducing more sophisticated MCMC algorithms. One of these algorithms is a mixture of two MCMC kernels: a random walk Metropolis kernel and a blockMetropolis-Hastings (MH) kernel with a variational approximation as proposaldistribution. The MH kernel allows one to locate regions of high probability efficiently. The Metropolis kernel allows us to explore the vicinity of these regions. This algorithm outperforms variationalapproximations because it yields slightly better estimates of the mean and considerably better estimates of higher moments, such as covariances. It also outperforms standard MCMC algorithms because it locates theregions of high probability quickly, thus speeding up convergence. We demonstrate this algorithm on the problem of Bayesian parameter estimation for logistic (sigmoid) belief networks.