Quantum control is concerned with the realisation of desired dynamics in quantum systems, serving as a linchpin for advancing quantum technologies and fundamental research. Analytic approaches and standard optimisation algorithms do not yield satisfactory solutions for more complex quantum systems, and especially not for real world quantum systems which are open and noisy. We devise a physics-constrained Reinforcement Learning (RL) algorithm that restricts the space of possible solutions. We incorporate priors about the desired time scales of the quantum state dynamics - as well as realistic control signal limitations - as constraints to the RL algorithm. These constraints improve solution quality and enhance computational scaleability. We evaluate our method on three broadly relevant quantum systems and incorporate real-world complications, arising from dissipation and control signal perturbations. We achieve both higher fidelities - which exceed 0.999 across all systems - and better robustness to time-dependent perturbations and experimental imperfections than previous methods. Lastly, we demonstrate that incorporating multi-step feedback can yield solutions robust even to strong perturbations. Our implementation can be found at: https://github.com/jan-o-e/RL4qcWpc.

Past work has studied the effects of fine-tuning on large language models' (LLMs) overall performance on certain tasks. However, a way to quantitatively and systematically analyze its effect on individual outputs is still lacking. In this work, we propose a new method for measuring the contribution that fine-tuning makes to individual LLM responses, assuming access to the original pre-trained model. Our method takes into account the model's intermediate hidden states, giving a more fine-grained insight into the effects of fine-tuning than a simple comparison of the final outputs of pre-trained and fine-tuned models. We introduce and theoretically analyze an exact decomposition of any fine-tuned LLM into a pre-training component and a fine-tuning component. Empirically, we find that one can steer model behavior and performance by up- or down-scaling the fine-tuning component during the forward pass. Motivated by this finding and our theoretical analysis, we define the Tuning Contribution ($\mathrm{TuCo}$) in terms of the ratio of the magnitudes fine-tuning component and the pre-training component. We find that three prominent adversarial attacks on LLMs circumvent safety measures in a way that reduces the Tuning Contribution, and that $\mathrm{TuCo}$ is consistently lower on prompts where the attacks succeed compared to ones where they do not. This suggests that attenuating the effect of fine-tuning on model outputs plays a role in the success of these attacks. In short, $\mathrm{TuCo}$ enables the quantitative study of how fine-tuning influences model behavior and safety, and vice-versa.

AI agents are commonly trained with large datasets of demonstrations of human behavior. However, not all behaviors are equally safe or desirable. Desired characteristics for an AI agent can be expressed by assigning desirability scores, which we assume are not assigned to individual behaviors but to collective trajectories. For example, in a dataset of vehicle interactions, these scores might relate to the number of incidents that occurred. We first assess the effect of each individual agent's behavior on the collective desirability score, e.g., assessing how likely an agent is to cause incidents. This allows us to selectively imitate agents with a positive effect, e.g., only imitating agents that are unlikely to cause incidents. To enable this, we propose the concept of an agent's \textit{Exchange Value}, which quantifies an individual agent's contribution to the collective desirability score. The Exchange Value is the expected change in desirability score when substituting the agent for a randomly selected agent. We propose additional methods for estimating Exchange Values from real-world datasets, enabling us to learn desired imitation policies that outperform relevant baselines. The project website can be found at https://tinyurl.com/select-to-perfect.

Autonomous agents deployed in the real world need to be robust against adversarial attacks on sensory inputs. Robustifying agent policies requires anticipating the strongest attacks possible. We demonstrate that existing observation-space attacks on reinforcement learning agents have a common weakness: while effective, their lack of information-theoretic detectability constraints makes them \textit{detectable} using automated means or human inspection. Detectability is undesirable to adversaries as it may trigger security escalations. We introduce \textit{\eattacks{}}, a novel form of adversarial attack on sequential decision-makers that is both effective and of $\epsilon-$bounded statistical detectability. We propose a novel dual ascent algorithm to learn such attacks end-to-end. Compared to existing attacks, we empirically find \eattacks{} to be significantly harder to detect with automated methods, and a small study with human participants\footnote{IRB approval under reference R84123/RE001} suggests they are similarly harder to detect for humans. Our findings suggest the need for better anomaly detectors, as well as effective hardware- and system-level defenses. The project website can be found at https://tinyurl.com/illusory-attacks.

Diffusion models have achieved remarkable results in image generation, and have similarly been used to learn high-performing policies in sequential decision-making tasks. Decision-making diffusion models can be trained on lower-quality data, and then be steered with a reward function to generate near-optimal trajectories. We consider the problem of extracting a reward function by comparing a decision-making diffusion model that models low-reward behavior and one that models high-reward behavior; a setting related to inverse reinforcement learning. We first define the notion of a \emph{relative reward function of two diffusion models} and show conditions under which it exists and is unique. We then devise a practical learning algorithm for extracting it by aligning the gradients of a reward function -- parametrized by a neural network -- to the difference in outputs of both diffusion models. Our method finds correct reward functions in navigation environments, and we demonstrate that steering the base model with the learned reward functions results in significantly increased performance in standard locomotion benchmarks. Finally, we demonstrate that our approach generalizes beyond sequential decision-making by learning a reward-like function from two large-scale image generation diffusion models. The extracted reward function successfully assigns lower rewards to harmful images.

We study how an autonomous agent learns to perform a task from demonstrations in a different domain, such as a different environment or different agent. Such cross-domain imitation learning is required to, for example, train an artificial agent from demonstrations of a human expert. We propose a scalable framework that enables cross-domain imitation learning without access to additional demonstrations or further domain knowledge. We jointly train the learner agent's policy and learn a mapping between the learner and expert domains with adversarial training. We effect this by using a mutual information criterion to find an embedding of the expert's state space that contains task-relevant information and is invariant to domain specifics. This step significantly simplifies estimating the mapping between the learner and expert domains and hence facilitates end-to-end learning. We demonstrate successful transfer of policies between considerably different domains, without extra supervision such as additional demonstrations, and in situations where other methods fail.

As connectomic datasets exceed hundreds of terabytes in size, accurate and efficient skeleton generation of the label volumes has evolved into a critical component of the computation pipeline used for analysis, evaluation, visualization, and error correction. We propose a novel topological thinning strategy that uses biological-constraints to produce accurate centerlines from segmented neuronal volumes while still maintaining biologically relevant properties. Current methods are either agnostic to the underlying biology, have non-linear running times as a function of the number of input voxels, or both. First, we eliminate from the input segmentation biologically-infeasible bubbles, pockets of voxels incorrectly labeled within a neuron, to improve segmentation accuracy, allow for more accurate centerlines, and increase processing speed. Next, a Convolutional Neural Network (CNN) detects cell bodies from the input segmentation, allowing us to anchor our skeletons to the somata. Lastly, a synapse-aware topological thinning approach produces expressive skeletons for each neuron with a nearly one-to-one correspondence between endpoints and synapses. We simultaneously estimate geometric properties of neurite width and geodesic distance between synapse and cell body, improving accuracy by 47.5% and 62.8% over baseline methods. We separate the skeletonization process into a series of computation steps, leveraging data-parallel strategies to increase throughput significantly. We demonstrate our results on over 1250 neurons and neuron fragments from three different species, processing over one million voxels per second per CPU with linear scalability.