Search

It has long been hypothesised that causal reasoning plays a fundamental role in robust and general intelligence. However, it is not known if agents must learn causal models in order to generalise to new domains, or if other inductive biases are sufficient. We answer this question, showing that any agent capable of satisfying a regret bound under a large set of distributional shifts must have learned an approximate causal model of the data generating process, which converges to the true causal model for optimal agents. We discuss the implications of this result for several research areas including transfer learning and causal inference., Comment: ICLR 2024 (oral). Updated agents section, new corollary

Causal models of agents have been used to analyse the safety aspects of machine learning systems. But identifying agents is non-trivial -- often the causal model is just assumed by the modeler without much justification -- and modelling failures can lead to mistakes in the safety analysis. This paper proposes the first formal causal definition of agents -- roughly that agents are systems that would adapt their policy if their actions influenced the world in a different way. From this we derive the first causal discovery algorithm for discovering agents from empirical data, and give algorithms for translating between causal models and game-theoretic influence diagrams. We demonstrate our approach by resolving some previous confusions caused by incorrect causal modelling of agents., Comment: Some typos corrected

To act safely and ethically in the real world, agents must be able to reason about harm and avoid harmful actions. However, to date there is no statistical method for measuring harm and factoring it into algorithmic decisions. In this paper we propose the first formal definition of harm and benefit using causal models. We show that any factual definition of harm must violate basic intuitions in certain scenarios, and show that standard machine learning algorithms that cannot perform counterfactual reasoning are guaranteed to pursue harmful policies following distributional shifts. We use our definition of harm to devise a framework for harm-averse decision making using counterfactual objective functions. We demonstrate this framework on the problem of identifying optimal drug doses using a dose-response model learned from randomized control trial data. We find that the standard method of selecting doses using treatment effects results in unnecessarily harmful doses, while our counterfactual approach allows us to identify doses that are significantly less harmful without sacrificing efficacy., Comment: Accepted at NeurIPS 2022. Included appendix comparing to Beckers et. al. arXiv:2210.05327. Typos corrected

Medical Triage is of paramount importance to healthcare systems, allowing for the correct orientation of patients and allocation of the necessary resources to treat them adequately. While reliable decision-tree methods exist to triage patients based on their presentation, those trees implicitly require human inference and are not immediately applicable in a fully automated setting. On the other hand, learning triage policies directly from experts may correct for some of the limitations of hard-coded decision-trees. In this work, we present a Deep Reinforcement Learning approach (a variant of DeepQ-Learning) to triage patients using curated clinical vignettes. The dataset, consisting of 1374 clinical vignettes, was created by medical doctors to represent real-life cases. Each vignette is associated with an average of 3.8 expert triage decisions given by medical doctors relying solely on medical history. We show that this approach is on a par with human performance, yielding safe triage decisions in 94% of cases, and matching expert decisions in 85% of cases. The trained agent learns when to stop asking questions, acquires optimized decision policies requiring less evidence than supervised approaches, and adapts to the novelty of a situation by asking for more information. Overall, we demonstrate that a Deep Reinforcement Learning approach can learn effective medical triage policies directly from expert decisions, without requiring expert knowledge engineering. This approach is scalable and can be deployed in healthcare settings or geographical regions with distinct triage specifications, or where trained experts are scarce, to improve decision making in the early stage of care., Comment: 17 pages, 4 figures, 3 tables, preprint, in press

The discovery of causal relationships is a fundamental problem in science and medicine. In recent years, many elegant approaches to discovering causal relationships between two variables from observational data have been proposed. However, most of these deal only with purely directed causal relationships and cannot detect latent common causes. Here, we devise a general heuristic which takes a causal discovery algorithm that can only distinguish purely directed causal relations and modifies it to also detect latent common causes. We apply our method to two directed causal discovery algorithms, the Information Geometric Causal Inference of (Daniusis et al., 2010) and the Kernel Conditional Deviance for Causal Inference of (Mitrovic, Sejdinovic, & Teh, 2018), and extensively test on synthetic data -- detecting latent common causes in additive, multiplicative and complex noise regimes -- and on real data, where we are able to detect known common causes. In addition to detecting latent common causes, our experiments demonstrate that both the modified algorithms preserve the performance of the original in distinguishing directed causal relations., Comment: Presented at Frontiers of AI-Assisted Care 2019 (Scientific Symposium, Stanford Medicine). 13 pages, 5 figures, 4 tables. Comments welcome

We elaborate on using importance sampling for causal reasoning, in particular for counterfactual inference. We show how this can be implemented natively in probabilistic programming. By considering the structure of the counterfactual query, one can significantly optimise the inference process. We also consider design choices to enable further optimisations. We introduce MultiVerse, a probabilistic programming prototype engine for approximate causal reasoning. We provide experimental results and compare with Pyro, an existing probabilistic programming framework with some of causal reasoning tools., Comment: Logan and Yura have made equal contributions to the paper. Accepted to the 2nd Symposium on Advances in Approximate Bayesian Inference (Vancouver, Canada, 2019)

Machine learning promises to revolutionize clinical decision making and diagnosis. In medical diagnosis a doctor aims to explain a patient's symptoms by determining the diseases \emph{causing} them. However, existing diagnostic algorithms are purely associative, identifying diseases that are strongly correlated with a patients symptoms and medical history. We show that this inability to disentangle correlation from causation can result in sub-optimal or dangerous diagnoses. To overcome this, we reformulate diagnosis as a counterfactual inference task and derive new counterfactual diagnostic algorithms. We show that this approach is closer to the diagnostic reasoning of clinicians and significantly improves the accuracy and safety of the resulting diagnoses. We compare our counterfactual algorithm to the standard Bayesian diagnostic algorithm and a cohort of 44 doctors using a test set of clinical vignettes. While the Bayesian algorithm achieves an accuracy comparable to the average doctor, placing in the top 48% of doctors in our cohort, our counterfactual algorithm places in the top 25% of doctors, achieving expert clinical accuracy. This improvement is achieved simply by changing how we query our model, without requiring any additional model improvements. Our results show that counterfactual reasoning is a vital missing ingredient for applying machine learning to medical diagnosis., Comment: Restructured and new subsections. Improved figures. Introduction rewritten

One of the most striking features of quantum theory is the existence of entangled states, responsible for Einstein's so called "spooky action at a distance". These states emerge from the mathematical formalism of quantum theory, but to date we do not have a clear idea of the physical principles that give rise to entanglement. Why does nature have entangled states? Would any theory superseding classical theory have entangled states, or is quantum theory special? One important feature of quantum theory is that it has a classical limit, recovering classical theory through the process of decoherence. We show that any theory with a classical limit must contain entangled states, thus establishing entanglement as an inevitable feature of any theory superseding classical theory.

The second law of thermodynamics states that a system in contact with a heat bath can undergo a transformation if and only if its free energy decreases. However, the "if" part of this statement is only true when the effective heat bath is infinite. In this article we remove this idealization and derive corrections to the second law in the case where the bath has a finite size, or equivalently finite heat capacity. This can also be translated to processes lasting a finite time, and we show that thermodynamical reversibility is lost in this regime. We do so in full generality, that is without assuming any particular model for the bath, the only parameters defining the bath are its temperature and heat capacity. We find connections with second order Shannon information theory, in particular in the case of Landauer erasure. We also consider the case of non-fluctuating work, and derive finite-bath corrections to the min and max free energies employed in single-shot thermodynamics., Comment: New references. Included discussion of finite size and more detailed comparison with existing results

One of the most striking features of quantum theory is the existence of entangled states, responsible for Einstein's so called "spooky action at a distance". These states emerge from the mathematical formalism of quantum theory, but to date we do not have a clear idea of which physical principles give rise to entanglement. Why does quantum theory have entangled states? Would any theory superseding classical theory have entangled states, or is quantum theory special? We demonstrate that without entanglement, non-classical degrees of freedom cannot reversibly interact. We present two postulates, no-cloning / no-broadcasting and local transitivity, either of which are sufficient to imply the existence of entangled states in any non-classical theory with reversible interactions. Therefore we argue that entanglement is an inevitable feature of a non-classical universe., Comment: 7 pages including appendices

In the standard framework of thermodynamics the work produced or consumed in a process is a random variable whose average value is bounded by the change in the free energy of the system. This work is calculated without regard for the size of its fluctuations. We find that in some processes, such as reversible cooling, the fluctuations of the work can diverge. Small or fragile thermal machines may be unable to cope with large fluctuations. Hence, with the present focus on nano- scale thermodynamics, we analyse how thermodynamic efficiency rates are modified when the size of the fluctuations around the average is restricted. We quantify the work content and work of formation of any state when the work fluctuations are bounded by a given amount c. By varying c we interpolate between the standard free energy c = infinity and the min-free energy c = 0, defined in the context of single-shot thermodynamics. We derive fundamental relations between average work and its fluctuations, and explore the emergence of irreversibility and partial order on state transformations when bounding fluctuations. We also study the efficiency a single qubit thermal engine model with constrained fluctuations, and derive the corrections to the Carnot efficiency., Comment: Published in Nature Communications

The richness of quantum theory's reversible dynamics is one of its unique operational characteristics, with recent results suggesting deep links between the theory's reversible dynamics, its local state space and the degree of non-locality it permits. We explore the delicate interplay between these features, demonstrating that reversibility places strong constraints on both the local and global state space. Firstly, we show that reversible dynamics are trivial (composed of local transformations and permutations of subsytems) in maximally non-local theories whose local state spaces satisfy a dichotomy criterion; this applies to a range of operational models that have previously been studied, such as d-dimensional "hyperballs" and almost all regular polytope systems. By separately deriving a similar result for odd-sided polygons, we show that classical systems are the only regular polytope state spaces whose maximally non-local composites allow for non-trivial reversible dynamics. Secondly, we show that non-trivial reversible dynamics do exist in maximally non-local theories whose state spaces are reducible into two or more smaller spaces. We conjecture that this is a necessary condition for the existence of such dynamics, but that reversible entanglement generation remains impossible even in this scenario., Comment: 12+epsilon pages, 2 figures

The standards in this document are for Hospitality and Tourism programs and are designed to clearly state what the student should know and be able to do upon completion of an advanced high-school program. Minimally, the student will complete a two-year program to achieve all standards. The Hospitality and Tourism Standards Writing Team followed the nationally-recognized certification programs established by the American Hotel and Lodging Association. The standards are organized as follows: Content Standards are general statements that identify major areas of knowledge, understanding and the skills students are expected to learn in key subject and career areas by the end of the program. Following each content standard are a number of performance standards. Performance standards identify the more specific components of each content standard and define the expected abilities of students within each content standard. Each performance standard is supported by performance indicators. Performance indicators are very specific criteria statements for determining whether a student exceeds the standard, meets the standard, or whose performance approaches the standard. All performance indicators are organized into three categories according to degree of difficulty: Exceeds Standard; Meets Standard; or Approaches Standard. Any student considered successful in the completion of a particular performance standard will have successfully demonstrated all performance indicators in the "Meets" and "Approaches" sections for the performance standard. Furthermore, any knowledge and/or performance assessments should be based on the indicators in those sections. Content standard 3.0 addresses employability skills, where students must demonstrate proficiency in workplace readiness, career development, and lifelong learning. The competencies in this section are centered around the 21st Century Skills initiative and designed to be completed by the student over the course of his/her high school career. The final section of the document shows where the performance standards and performance indicators support the state academic standards for math, science, and English Language Arts (ELA). Where correlation with an academic standard exists, students in the hospitality and tourism program perform learning activities that support, either directly or indirectly, achievement of one or more academic standards.

We study the long-range quantum correlations in the anisotropic XY-model. By first examining the thermodynamic limit we show that employing the quantum discord as a figure of merit allows one to capture the main features of the model at zero temperature. Further, by considering suitably large site separations we find that these correlations obey a simple scaling behavior for finite temperatures, allowing for efficient estimation of the critical point. We also address ground-state factorization of this model by explicitly considering finite size systems, showing its relation to the energy spectrum and explaining the persistence of the phenomenon at finite temperatures. Finally, we compute the fidelity between finite and infinite systems in order to show that remarkably small system sizes can closely approximate the thermodynamic limit., Comment: 8 pages, 8 figures. Close to published version