54 accepted papers
Previous work shows that the chain of continuous thought (continuous CoT) improves the reasoning capability of large language models (LLMs) by enabling implicit parallel thinking, and a subsequent work provided theoretical insight by showing that a two-layer transformer equipped with continuous CoT…
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 dat…
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 g…
Most learning algorithms with formal regret guarantees assume that all mistakes are recoverable and essentially rely on trying all possible behaviors. This approach is problematic when some mistakes are *catastrophic*, i.e., irreparable. We propose an online learning problem where the goal is to min…
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 anticip…
Large language models generate code one token at a time. Their autoregressive generation process lacks the feedback of observing the program's output. Training LLMs to suggest edits directly can be challenging due to the scarcity of rich edit data. To address these problems, we propose neural diffus…
Pretrained language models (LMs) can generalize to implications of facts that they are finetuned on. For example, if finetuned on "John Doe lives in Tokyo," LMs correctly answer "What language do the people in John Doe's city speak?'' with "Japanese''. However, little is known about the mechanisms t…
Large language models (LLMs) can acquire new knowledge through fine-tuning, but this process exhibits a puzzling duality: models can generalize remarkably from new facts, yet are also prone to hallucinating incorrect information. However, the reasons for this phenomenon remain poorly understood. In…
Language models (LMs) are susceptible to bias, sycophancy, backdoors, and other tendencies that lead to unfaithful responses to the input context. Interpreting internal states of LMs could help monitor and correct unfaithful behavior. We hypothesize that LMs faithfully represent their input contexts…
We study partially observable assistance games (POAGs), a model of the human-AI value alignment problem which allows the human and the AI assistant to have partial observations. Motivated by concerns of AI deception, we study a qualitatively new phenomenon made possible by partial observability: wou…
In reinforcement learning (RL), if the agent’s reward differs from the designers’ true utility, even only rarely, the state distribution resulting from the agent’s policy can be very bad, in theory and in practice. When RL policies would devolve into undesired behavior, a common countermeasure is KL
Large Language Models (LLMs) have demonstrated remarkable performance in many applications, including challenging reasoning problems via chain-of-thought (CoT) techniques that generate ``thinking tokens'' before answering the questions. While existing theoretical works demonstrate that CoT with disc…
Adversarial optimization algorithms that explicitly search for flaws in agents' policies have been successfully applied to finding robust and diverse policies in the context of multi-agent learning. However, the success of adversarial optimization has been largely limited to zero-sum settings becaus…
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.'' 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…
Existing AI alignment approaches assume that preferences are static, which is unrealistic: our preferences change, and may even be influenced by our interactions with AI systems themselves. To clarify the consequences of incorrectly assuming static preferences, we introduce Dynamic Reward Markov Dec…
Ethically compliant autonomous systems (ECAS) are the prevailing approach to building robotic systems that perform sequential decision making subject to ethical theories in fully observable environments. However, in real-world robotics settings, these systems often operate under partial observabilit…
Do neural networks learn to implement algorithms such as look-ahead or search "in the wild"? Or do they rely purely on collections of simple heuristics? We present evidence of *learned look-ahead* in the policy and value network of Leela Chess Zero, the currently strongest deep neural chess engine.…
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 i…
We study the representation complexity of model-based and model-free reinforcement learning (RL) in the context of circuit complexity. We prove theoretically that there exists a broad class of MDPs such that their underlying transition and reward functions can be represented by constant depth circui…
Foundation models such as GPT-4 are fine-tuned to avoid unsafe or otherwise problematic behavior, such as helping to commit crimes or producing racist text. One approach to fine-tuning, called reinforcement learning from human feedback, learns from humans’ expressed preferences over multiple outputs…
While Large Language Models (LLMs) are increasingly being used in real-world applications, they remain vulnerable to *prompt injection attacks*: malicious third party prompts that subvert the intent of the system designer. To help researchers study this problem, we present a dataset of over 563,000…
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 w…
Auto-regressive large language models (LLMs) show impressive capacities to solve many complex reasoning tasks while struggling with some simple logical reasoning tasks such as inverse search: when trained on ''$A \to B$'' (e.g., *Tom is the parent of John*), LLM fails to directly conclude ''$B \gets…
Learning from human feedback has gained traction in fields like robotics and natural language processing in recent years. While prior works mostly rely on human feedback in the form of comparisons, language is a preferable modality that provides more informative insights into user preferences. In th…
Past analyses of reinforcement learning from human feedback (RLHF) assume that the human evaluators fully observe the environment. What happens when human feedback is based only on partial observations? We formally define two failure cases: deceptive inflation and overjustification. Modeling the hum…
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 transfer…
Deep reinforcement learning (RL) works impressively in some environments and fails catastrophically in others. Ideally, RL theory should be able to provide an understanding of why this is, i.e. bounds predictive of practical performance. Unfortunately, current theory does not quite have this ability…
Robotic systems are often composed of modular algorithms that each perform a specific function within a larger architecture, ranging from state estimation and task planning to trajectory optimization and object recognition. Existing work for specifying these systems as a formal composition of contra…
It is often very challenging to manually design reward functions for complex, real-world tasks. To solve this, one can instead use reward learning to infer a reward function from data. However, there are often multiple reward functions that fit the data equally well, even in the infinite-data limit.…
Offline reinforcement learning (RL), which aims at learning good policies from historical data, 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 con…
This paper introduces SMCP3, a method for automatically implementing custom sequential Monte Carlo samplers for inference in probabilistic programs. Unlike particle filters and resample-move SMC (Gilks and Berzuini, 2001), SMCP3 algorithms can improve the quality of samples and weights using pairs o…
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 accelerat…
Cross-domain imitation learning studies how to leverage expert demonstrations of one agent to train an imitation agent with a different embodiment or morphology. Comparing trajectories and stationary distributions between the expert and imitation agents is challenging because they live on different…
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…
Although it has been known since the 1970s that a
Markov decision processes (MDPs) are a common general-purpose model used in robotics for representing sequential decision-making problems. Given the complexity of robotics applications, a popular approach for approximately solving MDPs relies on state aggregation to reduce the size of the state spac…
Imitation learning often needs a large demonstration set in order to handle the full range of situations that an agent might find itself in during deployment. However, collecting expert demonstrations can be expensive. Recent work in vision, reinforcement learning, and NLP has shown that auxiliary r…
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 methods are used: imitation learning which is suitable for expert datasets, and vanilla offline RL…
In online reinforcement learning (RL), efficient exploration remains particularly challenging in high-dimensional environments with sparse rewards. In low-dimensional environments, where tabular parameterization is possible, count-based upper confidence bound (UCB) exploration methods achieve minima…
For many tasks, the reward function is inaccessible to introspection or too complex to be specified procedurally, and must instead be learned from user data. Prior work has evaluated learned reward functions by evaluating policies optimized for the learned reward. However, this method cannot disting…
Lookahead search has been a critical component of recent AI successes, such as in the games of chess, go, and poker. However, the search methods used in these games, and in many other settings, are tabular. Tabular search methods do not scale well with the size of the search space, and this problem…
Existing observational approaches for learning human preferences, such as inverse reinforcement learning, usually make strong assumptions about the observability of the human's environment. However, in reality, people make many important decisions under uncertainty. To better understand preference l…
Deep reinforcement learning (RL) policies are known to be vulnerable to adversarial perturbations to their observations, similar to adversarial examples for classifiers. However, an attacker is not usually able to directly modify another agent's observations. This might lead one to wonder: is it pos…
We introduce a new memory architecture, Bayesian Relational Memory (BRM), to improve the generalization ability for semantic visual navigation agents in unseen environments, where an agent is given a semantic target to navigate towards. BRM takes the form of a probabilistic relation graph over seman…
Our goal is for AI systems to correctly identify and act according to their human user’s objectives. Cooperative Inverse Reinforcement Learning (CIRL) formalizes this value alignment problem as a two-player game between a human and robot, in which only the human knows the parameters of the reward fu…
Despite the recent successes of probabilistic programming languages (PPLs) in AI applications, PPLs offer only limited support for random variables whose distributions combine discrete and continuous elements. We develop the notion of measure-theoretic Bayesian networks (MTBNs) and use it to provide…
In recent years, deep generative models have been shown to 'imagine' convincing high-dimensional observations such as images, audio, and even video, learning directly from raw data. In this work, we ask how to imagine goal-directed visual plans -- a plausible sequence of observations that transition…
Effective implementations of sampling-based probabilistic inference often require manually constructed, model-specific proposals. Inspired by recent progresses in meta-learning for training learning agents that can generalize to unseen environments, we propose a meta-learning approach to building ef…
It is commonly believed that an agent making decisions on behalf of two or more principals who have different utility functions should adopt a Pareto optimal policy, i.e. a policy that cannot be improved upon for one principal without making sacrifices for another. Harsanyi's theorem shows that when…
Autonomous agents optimize the reward function we give them. What they don't know is how hard it is for us to design a reward function that actually captures what we want. When designing the reward, we might think of some specific training scenarios, and make sure that the reward will lead to the r…
Detecting weak seismic events from noisy sensors is a difficult perceptual task. We formulate this task as Bayesian inference and propose a generative model of seismic events and signals across a network of spatially distributed stations. Our system, SIGVISA, is the first to directly model seismic w…
For an autonomous system to be helpful to humans and to pose no unwarranted risks, it needs to align its values with those of the humans in its environment in such a way that its actions contribute to the maximization of value for the humans. We propose a formal definition of the value alignment pro…
We consider the problem of refining an abstract task plan into a motion trajectory. Task and motion planning is a hard problem that is essential to long-horizon mobile manipulation. Many approaches divide the problem into two steps: a search for a task plan and task plan refinement to find a feasibl…
Gaussian processes have been successful in both supervised and unsupervised machine learning tasks, but their computational complexity has constrained practical applications. We introduce a new approximation for large-scale Gaussian processes, the Gaussian Process Random Field (GPRF), in which local…