Search

Matching is a commonly used causal inference framework in observational studies. By pairing individuals with different treatment values but with the same values of covariates (i.e., exact matching), the sample average treatment effect (SATE) can be consistently estimated and inferred using the classic Neyman-type (difference-in-means) estimator and confidence interval. However, inexact matching typically exists in practice and may cause substantial bias for the downstream treatment effect estimation and inference. Many methods have been proposed to reduce bias due to inexact matching in the binary treatment case. However, to our knowledge, no existing work has systematically investigated bias due to inexact matching in the continuous treatment case. To fill this blank, we propose a general framework for reducing bias in inexactly matched observational studies with continuous treatments. In the matching stage, we propose a carefully formulated caliper that incorporates the information of both the paired covariates and treatment doses to better tailor matching for the downstream SATE estimation and inference. In the estimation and inference stage, we propose a bias-corrected Neyman estimator paired with the corresponding bias-corrected variance estimator to leverage the information on propensity density discrepancies after inexact matching to further reduce the bias due to inexact matching. We apply our proposed framework to COVID-19 social mobility data to showcase differences between classic and bias-corrected SATE estimation and inference.

We introduce a class of topological pairing orders characterized by a half-integer pair monopole charge, leading to Berry phase enforced half-integer partial wave symmetry. This exotic spinor order emerges from pairing between Fermi surfaces with Chern numbers differing by an odd integer. Using tight-binding models, we demonstrate spinor superconducting orders with monopole charges $\pm 1/2$, featuring a single gap node and nontrivial surface states. Additionally, the superfluid velocity follows a fractionalized Mermin-Ho relation in spatially inhomogeneous pairing orders. The concept extends to spinor density waves and excitons., Comment: This article supersedes arXiv:2001.05984

We analyse the impact of using tempered likelihoods in the production of posterior predictions. Our findings reveal that once the sample size is at least moderately large and the temperature is not too small, then likelihood tempering has virtually no impact on the resulting posterior predictions., Comment: 7 (+14 supplementary) pages (excluding references); 3 (+4 supplementary) figures

Counterexample-driven genetic programming (CDGP) uses specifications provided as formal constraints to generate the training cases used to evaluate evolving programs. It has also been extended to combine formal constraints and user-provided training data to solve symbolic regression problems. Here we show how the ideas underlying CDGP can also be applied using only user-provided training data, without formal specifications. We demonstrate the application of this method, called ``informal CDGP,'' to software synthesis problems. Our results show that informal CDGP finds solutions faster (i.e. with fewer program executions) than standard GP. Additionally, we propose two new variants to informal CDGP, and find that one produces significantly more successful runs on about half of the tested problems. Finally, we study whether the addition of counterexample training cases to the training set is useful by comparing informal CDGP to using a static subsample of the training set, and find that the addition of counterexamples significantly improves performance.

Distinct from familiar $s$-, $p$-, or $d$-wave pairings, the monopole superconducting order represents a novel class of pairing order arising from nontrivial monopole charge of the Cooper pair. In the weak-coupling regime, this order can emerge when pairing occurs between Fermi surfaces with different Chern numbers in, for example, doped Weyl semimetal systems. However, the phase of monopole pairing order is not well-defined over an entire Fermi surface, making it challenging to design experiments sensitive to both its symmetry and topology. To address this, we propose a scheme based on symmetry and topological principles to identify this elusive pairing order through a set of phase-sensitive Josephson experiments. By examining the discrepancy between global and local angular momentum of the pairing order, we can unveil the monopole charge of the pairing order. We demonstrate the proposed probe of monopole pairing order through analytic and numerical studies of Josephson coupling in models of monopole superconductor junctions. This work opens a promising avenue to uncover the unique topological properties of monopole pairing orders and to distinguish them from known pairing orders based on spherical harmonic symmetry.

Bayesian optimization is a technique for efficiently optimizing unknown functions in a black-box manner. To handle practical settings where gathering data requires use of finite resources, it is desirable to explicitly incorporate function evaluation costs into Bayesian optimization policies. To understand how to do so, we develop a previously-unexplored connection between cost-aware Bayesian optimization and the Pandora's Box problem, a decision problem from economics. The Pandora's Box problem admits a Bayesian-optimal solution based on an expression called the Gittins index, which can be reinterpreted as an acquisition function. We study the use of this acquisition function for cost-aware Bayesian optimization, and demonstrate empirically that it performs well, particularly in medium-high dimensions. We further show that this performance carries over to classical Bayesian optimization without explicit evaluation costs. Our work constitutes a first step towards integrating techniques from Gittins index theory into Bayesian optimization.

We consider the predict-then-optimize paradigm for decision-making in which a practitioner (1) trains a supervised learning model on historical data of decisions, contexts, and rewards, and then (2) uses the resulting model to make future binary decisions for new contexts by finding the decision that maximizes the model's predicted reward. This approach is common in industry. Past analysis assumes that rewards are observed for all actions for all historical contexts, which is possible only in problems with special structure. Motivated by problems from ads targeting and recommender systems, we study new black-box predict-then-optimize problems that lack this special structure and where we only observe the reward from the action taken. We present a novel loss function, which we call Empirical Soft Regret (ESR), designed to significantly improve reward when used in training compared to classical accuracy-based metrics like mean-squared error. This loss function targets the regret achieved when taking a suboptimal decision; because the regret is generally not differentiable, we propose a differentiable "soft" regret term that allows the use of neural networks and other flexible machine learning models dependent on gradient-based training. In the particular case of paired data, we show theoretically that optimizing our loss function yields asymptotically optimal regret within the class of supervised learning models. We also show our approach significantly outperforms state-of-the-art algorithms on real-world decision-making problems in news recommendation and personalized healthcare compared to benchmark methods from contextual bandits and conditional average treatment effect estimation., Comment: 15 pages, 2 figures, 3 tables

Large Language Models (LLMs) have demonstrated potential in cybersecurity applications but have also caused lower confidence due to problems like hallucinations and a lack of truthfulness. Existing benchmarks provide general evaluations but do not sufficiently address the practical and applied aspects of LLM performance in cybersecurity-specific tasks. To address this gap, we introduce the SECURE (Security Extraction, Understanding \& Reasoning Evaluation), a benchmark designed to assess LLMs performance in realistic cybersecurity scenarios. SECURE includes six datasets focussed on the Industrial Control System sector to evaluate knowledge extraction, understanding, and reasoning based on industry-standard sources. Our study evaluates seven state-of-the-art models on these tasks, providing insights into their strengths and weaknesses in cybersecurity contexts, and offer recommendations for improving LLMs reliability as cyber advisory tools.

In recent years, the shortcomings of Bayes posteriors as inferential devices has received increased attention. A popular strategy for fixing them has been to instead target a Gibbs measure based on losses that connect a parameter of interest to observed data. While existing theory for such inference procedures relies on these losses to be analytically available, in many situations these losses must be stochastically estimated using pseudo-observations. The current paper fills this research gap, and derives the first asymptotic theory for Gibbs measures based on estimated losses. Our findings reveal that the number of pseudo-observations required to accurately approximate the exact Gibbs measure depends on the rates at which the bias and variance of the estimated loss converge to zero. These results are particularly consequential for the emerging field of generalised Bayesian inference, for estimated intractable likelihoods, and for biased pseudo-marginal approaches. We apply our results to three Gibbs measures that have been proposed to deal with intractable likelihoods and model misspecification.

Chemistry plays a crucial role in many domains, such as drug discovery and material science. While large language models (LLMs) such as GPT-4 exhibit remarkable capabilities on natural language processing tasks, existing research indicates that their performance on chemistry tasks is discouragingly low. In this paper, however, we demonstrate that our developed LLMs can achieve very strong results on a comprehensive set of chemistry tasks, outperforming the most advanced GPT-4 and Claude 3 Opus by a substantial margin. To accomplish this, we propose SMolInstruct, a large-scale, comprehensive, and high-quality dataset for instruction tuning. It contains 14 selected chemistry tasks and over three million samples, laying a solid foundation for training and evaluating LLMs for chemistry. Using SMolInstruct, we fine-tune a set of open-source LLMs, among which, we find that Mistral serves as the best base model for chemistry tasks. Our analysis further demonstrates the critical role of the proposed dataset in driving the performance improvements., Comment: Accepted by COLM 2024

Experimentation with interference poses a significant challenge in contemporary online platforms. Prior research on experimentation with interference has concentrated on the final output of a policy. The cumulative performance, while equally crucial, is less well understood. To address this gap, we introduce the problem of {\em Multi-armed Bandits with Interference} (MABI), where the learner assigns an arm to each of $N$ experimental units over a time horizon of $T$ rounds. The reward of each unit in each round depends on the treatments of {\em all} units, where the influence of a unit decays in the spatial distance between units. Furthermore, we employ a general setup wherein the reward functions are chosen by an adversary and may vary arbitrarily across rounds and units. We first show that switchback policies achieve an optimal {\em expected} regret $\tilde O(\sqrt T)$ against the best fixed-arm policy. Nonetheless, the regret (as a random variable) for any switchback policy suffers a high variance, as it does not account for $N$. We propose a cluster randomization policy whose regret (i) is optimal in {\em expectation} and (ii) admits a high probability bound that vanishes in $N$.

Bayesian optimization is a powerful framework for optimizing functions that are expensive or time-consuming to evaluate. Recent work has considered Bayesian optimization of function networks (BOFN), where the objective function is given by a network of functions, each taking as input the output of previous nodes in the network as well as additional parameters. Leveraging this network structure has been shown to yield significant performance improvements. Existing BOFN algorithms for general-purpose networks evaluate the full network at each iteration. However, many real-world applications allow for evaluating nodes individually. To exploit this, we propose a novel knowledge gradient acquisition function that chooses which node and corresponding inputs to evaluate in a cost-aware manner, thereby reducing query costs by evaluating only on a part of the network at each step. We provide an efficient approach to optimizing our acquisition function and show that it outperforms existing BOFN methods and other benchmarks across several synthetic and real-world problems. Our acquisition function is the first to enable cost-aware optimization of a broad class of function networks., Comment: 34 pages, 15 figures, 3 tables

Approximate Bayesian Computation (ABC) has gained popularity as a method for conducting inference and forecasting in complex models, most notably those which are intractable in some sense. In this paper we use ABC to produce probabilistic forecasts in state space models (SSMs). Whilst ABC-based forecasting in correctly-specified SSMs has been studied, the misspecified case has not been investigated, and it is that case which we emphasize. We invoke recent principles of 'focused' Bayesian prediction, whereby Bayesian updates are driven by a scoring rule that rewards predictive accuracy; the aim being to produce predictives that perform well in that rule, despite misspecification. Two methods are investigated for producing the focused predictions. In a simulation setting, 'coherent' predictions are in evidence for both methods: the predictive constructed via the use of a particular scoring rule predicts best according to that rule. Importantly, both focused methods typically produce more accurate forecasts than an exact, but misspecified, predictive. An empirical application to a truly intractable SSM completes the paper.

Given rapid progress toward advanced AI and risks from frontier AI systems (advanced AI systems pushing the boundaries of the AI capabilities frontier), the creation and implementation of AI governance and regulatory schemes deserves prioritization and substantial investment. However, the status quo is untenable and, frankly, dangerous. A regulatory gap has permitted AI labs to conduct research, development, and deployment activities with minimal oversight. In response, frontier AI system evaluations have been proposed as a way of assessing risks from the development and deployment of frontier AI systems. Yet, the budding AI risk evaluation ecosystem faces significant coordination challenges, such as a limited diversity of evaluators, suboptimal allocation of effort, and perverse incentives. This paper proposes a solution in the form of an international consortium for AI risk evaluations, comprising both AI developers and third-party AI risk evaluators. Such a consortium could play a critical role in international efforts to mitigate societal-scale risks from advanced AI, including in managing responsible scaling policies and coordinated evaluation-based risk response. In this paper, we discuss the current evaluation ecosystem and its shortcomings, propose an international consortium for advanced AI risk evaluations, discuss issues regarding its implementation, discuss lessons that can be learnt from previous international institutions and existing proposals for international AI governance institutions, and, finally, we recommend concrete steps to advance the establishment of the proposed consortium: (i) solicit feedback from stakeholders, (ii) conduct additional research, (iii) conduct a workshop(s) for stakeholders, (iv) analyze feedback and create final proposal, (v) solicit funding, and (vi) create a consortium., Comment: 50 pages, 2 figures; updated w/ a few minor revisions based on feedback from SoLaR Workshop reviewers (on 5 page version)

Simulation has become, in many application areas, a sine-qua-non. Most recently, COVID-19 has underlined the importance of simulation studies and limitations in current practices and methods. We identify four goals of methodological work for addressing these limitations. The first is to provide better support for capturing, representing, and evaluating the context of simulation studies, including research questions, assumptions, requirements, and activities contributing to a simulation study. In addition, the composition of simulation models and other simulation studies' products must be supported beyond syntactical coherence, including aspects of semantics and purpose, enabling their effective reuse. A higher degree of automating simulation studies will contribute to more systematic, standardized simulation studies and their efficiency. Finally, it is essential to invest increased effort into effectively communicating results and the processes involved in simulation studies to enable their use in research and decision-making. These goals are not pursued independently of each other, but they will benefit from and sometimes even rely on advances in other subfields. In the present paper, we explore the basis and interdependencies evident in current research and practice and delineate future research directions based on these considerations.

During the COVID-19 pandemic, safely implementing in-person indoor instruction was a high priority for universities nationwide. To support this effort at the University, we developed a mathematical model for estimating the risk of SARS-CoV-2 transmission in university classrooms. This model was used to evaluate combinations of feasible interventions for classrooms at the University during the pandemic and optimize the set of interventions that would allow higher occupancy levels, matching the pre-pandemic numbers of in-person courses. Importantly, we determined that requiring masking in dense classrooms with unrestricted seating with more than 90% of students vaccinated was easy to implement, incurred little logistical or financial cost, and allowed classes to be held at full capacity. A retrospective analysis at the end of the semester confirmed the model's assessment that the proposed classroom configuration would be safe. Our framework is generalizable and was used to support reopening decisions at Stanford University. In addition, our framework is flexible and applies to a wide range of indoor settings. It was repurposed for large university events and gatherings and could be used to support planning indoor space use to avoid transmission of infectious diseases across various industries, from secondary schools to movie theaters and restaurants.

In copula models the marginal distributions and copula function are specified separately. We treat these as two modules in a modular Bayesian inference framework, and propose conducting modified Bayesian inference by "cutting feedback". Cutting feedback limits the influence of potentially misspecified modules in posterior inference. We consider two types of cuts. The first limits the influence of a misspecified copula on inference for the marginals, which is a Bayesian analogue of the popular Inference for Margins (IFM) estimator. The second limits the influence of misspecified marginals on inference for the copula parameters by using a pseudo likelihood of the ranks to define the cut model. We establish that if only one of the modules is misspecified, then the appropriate cut posterior gives accurate uncertainty quantification asymptotically for the parameters in the other module. Computation of the cut posteriors is difficult, and new variational inference methods to do so are proposed. The efficacy of the new methodology is demonstrated using both simulated data and a substantive multivariate time series application from macroeconomic forecasting. In the latter, cutting feedback from misspecified marginals to a 1096 dimension copula improves posterior inference and predictive accuracy greatly, compared to conventional Bayesian inference.

Retrosynthesis is the process of determining the set of reactant molecules that can react to form a desired product. Semi-template-based retrosynthesis methods, which imitate the reverse logic of synthesis reactions, first predict the reaction centers in the products, and then complete the resulting synthons back into reactants. We develop a new offline-online reinforcement learning method RLSynC for synthon completion in semi-template-based methods. RLSynC assigns one agent to each synthon, all of which complete the synthons by conducting actions step by step in a synchronized fashion. RLSynC learns the policy from both offline training episodes and online interactions, which allows RLSynC to explore new reaction spaces. RLSynC uses a standalone forward synthesis model to evaluate the likelihood of the predicted reactants in synthesizing a product, and thus guides the action search. Our results demonstrate that RLSynC can outperform state-of-the-art synthon completion methods with improvements as high as 14.9%, highlighting its potential in synthesis planning., Comment: 32 pages, 5 figures, 4 tables

We demonstrate that the forecasting combination puzzle is a consequence of the methodology commonly used to produce forecast combinations. By the combination puzzle, we refer to the empirical finding that predictions formed by combining multiple forecasts in ways that seek to optimize forecast performance often do not out-perform more naive, e.g. equally-weighted, approaches. In particular, we demonstrate that, due to the manner in which such forecasts are typically produced, tests that aim to discriminate between the predictive accuracy of competing combination strategies can have low power, and can lack size control, leading to an outcome that favours the naive approach. We show that this poor performance is due to the behavior of the corresponding test statistic, which has a non-standard asymptotic distribution under the null hypothesis of no inferior predictive accuracy, rather than the {standard normal distribution that is} {typically adopted}. In addition, we demonstrate that the low power of such predictive accuracy tests in the forecast combination setting can be completely avoided if more efficient estimation strategies are used in the production of the combinations, when feasible. We illustrate these findings both in the context of forecasting a functional of interest and in terms of predictive densities. A short empirical example {using daily financial returns} exemplifies how researchers can avoid the puzzle in practical settings.