Search

In many social science applications, researchers use the difference-in-differences (DID) estimator to establish causal relationships, exploiting cross-sectional variation in a baseline factor and temporal variation in exposure to an event that presumably may affect all units. This approach, which we term factorial DID (FDID), differs from canonical DID in that it lacks a clean control group unexposed to the event after the event occurs. In this paper, we clarify FDID as a research design in terms of its data structure, feasible estimands, and identifying assumptions that allow the DID estimator to recover these estimands. We frame FDID as a factorial design with two factors: the baseline factor, denoted by $G$, and the exposure level to the event, denoted by $Z$, and define the effect modification and causal interaction as the associative and causal effects of $G$ on the effect of $Z$, respectively. We show that under the canonical no anticipation and parallel trends assumptions, the DID estimator identifies only the effect modification of $G$ in FDID, and propose an additional factorial parallel trends assumption to identify the causal interaction. Moreover, we show that the canonical DID research design can be reframed as a special case of the FDID research design with an additional exclusion restriction assumption, thereby reconciling the two approaches. We extend this framework to allow conditionally valid parallel trends assumptions and multiple time periods, and clarify assumptions required to justify regression analysis under FDID. We illustrate these findings with empirical examples from economics and political science, and provide recommendations for improving practice and interpretation under FDID.

In 1986, Robert LaLonde published an article that compared nonexperimental estimates to experimental benchmarks (LaLonde 1986). He concluded that the nonexperimental methods at the time could not systematically replicate experimental benchmarks, casting doubt on the credibility of these methods. Following LaLonde's critical assessment, there have been significant methodological advances and practical changes, including (i) an emphasis on estimators based on unconfoundedness, (ii) a focus on the importance of overlap in covariate distributions, (iii) the introduction of propensity score-based methods leading to doubly robust estimators, (iv) a greater emphasis on validation exercises to bolster research credibility, and (v) methods for estimating and exploiting treatment effect heterogeneity. To demonstrate the practical lessons from these advances, we reexamine the LaLonde data and the Imbens-Rubin-Sacerdote lottery data. We show that modern methods, when applied in contexts with sufficient covariate overlap, yield robust estimates for the adjusted differences between the treatment and control groups. However, this does not mean that these estimates are valid. To assess their credibility, validation exercises (such as placebo tests) are essential, whereas goodness of fit tests alone are inadequate. Our findings highlight the importance of closely examining the assignment process, carefully inspecting overlap, and conducting validation exercises when analyzing causal effects with nonexperimental data.

This paper studies the challenge of developing robots capable of understanding under-specified instructions for creating functional object arrangements, such as "set up a dining table for two"; previous arrangement approaches have focused on much more explicit instructions, such as "put object A on the table." We introduce a framework, SetItUp, for learning to interpret under-specified instructions. SetItUp takes a small number of training examples and a human-crafted program sketch to uncover arrangement rules for specific scene types. By leveraging an intermediate graph-like representation of abstract spatial relationships among objects, SetItUp decomposes the arrangement problem into two subproblems: i) learning the arrangement patterns from limited data and ii) grounding these abstract relationships into object poses. SetItUp leverages large language models (LLMs) to propose the abstract spatial relationships among objects in novel scenes as the constraints to be satisfied; then, it composes a library of diffusion models associated with these abstract relationships to find object poses that satisfy the constraints. We validate our framework on a dataset comprising study desks, dining tables, and coffee tables, with the results showing superior performance in generating physically plausible, functional, and aesthetically pleasing object arrangements compared to existing models., Comment: 10 pages main paper, 21 pages appendix, RSS 2024

Phase-coherent pulsed driving of passive optical fiber resonators enable the generation of ultrashort dissipative Raman solitons with durations well below 100~fs. The existence and characteristics of such solitons critically depends on the desynchronisation between the pulsed driving source and the resonator roundtrip time, yet the full mechanism through which these dependencies arise remains unclear. Here, we numerically demonstrate that Raman solitons can exist even under conditions of continuous wave driving, and by numerically examining the existence and characteristics of Raman solitons under such conditions, we elucidate the role of desynchronisation in pulse-driven systems. In addition to providing new insights on the existence and characteristics of ultrashort Raman solitons, our analysis yields a qualitative explanation for the range of desynchronisations over which the solitons can exist.

Deformable object manipulation is a long-standing challenge in robotics. While existing approaches often focus narrowly on a specific type of object, we seek a general-purpose algorithm, capable of manipulating many different types of objects: beans, rope, cloth, liquid, . . . . One key difficulty is a suitable representation, rich enough to capture object shape, dynamics for manipulation and yet simple enough to be acquired effectively from sensor data. Specifically, we propose Differentiable Particles (DiPac), a new algorithm for deformable object manipulation. DiPac represents a deformable object as a set of particles and uses a differentiable particle dynamics simulator to reason about robot manipulation. To find the best manipulation action, DiPac combines learning, planning, and trajectory optimization through differentiable trajectory tree optimization. Differentiable dynamics provides significant benefits and enable DiPac to (i) estimate the dynamics parameters efficiently, thereby narrowing the sim-to-real gap, and (ii) choose the best action by backpropagating the gradient along sampled trajectories. Both simulation and real-robot experiments show promising results. DiPac handles a variety of object types. By combining planning and learning, DiPac outperforms both pure model-based planning methods and pure data-driven learning methods. In addition, DiPac is robust and adapts to changes in dynamics, thereby enabling the transfer of an expert policy from one object to another with different physical properties, e.g., from a rigid rod to a deformable rope.

The advantages of pre-trained large language models (LLMs) are apparent in a variety of language processing tasks. But can a language model's knowledge be further harnessed to effectively disambiguate objects and navigate decision-making challenges within the realm of robotics? Our study reveals the LLM's aptitude for solving complex decision making challenges that are often previously modeled by Partially Observable Markov Decision Processes (POMDPs). A pivotal focus of our research is the object disambiguation capability of LLMs. We detail the integration of an LLM into a tabletop environment disambiguation task, a decision making problem where the robot's task is to discern and retrieve a user's desired object from an arbitrarily large and complex cluster of objects. Despite multiple query attempts with zero-shot prompt engineering (details can be found in the Appendix), the LLM struggled to inquire about features not explicitly provided in the scene description. In response, we have developed a few-shot prompt engineering system to improve the LLM's ability to pose disambiguating queries. The result is a model capable of both using given features when they are available and inferring new relevant features when necessary, to successfully generate and navigate down a precise decision tree to the correct object--even when faced with identical options.

Learning reward functions for physical skills are challenging due to the vast spectrum of skills, the high-dimensionality of state and action space, and nuanced sensory feedback. The complexity of these tasks makes acquiring expert demonstration data both costly and time-consuming. Large Language Models (LLMs) contain valuable task-related knowledge that can aid in learning these reward functions. However, the direct application of LLMs for proposing reward functions has its limitations such as numerical instability and inability to incorporate the environment feedback. We aim to extract task knowledge from LLMs using environment feedback to create efficient reward functions for physical skills. Our approach consists of two components. We first use the LLM to propose features and parameterization of the reward function. Next, we update the parameters of this proposed reward function through an iterative self-alignment process. In particular, this process minimizes the ranking inconsistency between the LLM and our learned reward functions based on the new observations. We validated our method by testing it on three simulated physical skill learning tasks, demonstrating effective support for our design choices., Comment: CoRL 2023, LangRob workshop

Two-way fixed effects (TWFE) models are ubiquitous in causal panel analysis in political science. However, recent methodological discussions challenge their validity in the presence of heterogeneous treatment effects (HTE) and violations of the parallel trends assumption (PTA). This burgeoning literature has introduced multiple estimators and diagnostics, leading to confusion among empirical researchers on two fronts: the reliability of existing results based on TWFE models and the current best practices. To address these concerns, we examined, replicated, and reanalyzed 37 articles from three leading political science journals that employed observational panel data with binary treatments. Using six newly introduced HTE-robust estimators, along with diagnostics tests and uncertainty measures that are robust to PTA violations, we find that only a small minority of studies are highly robust. Although HTE-robust estimates tend to be broadly consistent with TWFE estimates, discrepancies in point estimates, increased measures of uncertainty, and potential PTA violations call into question many results that were already on the margins of statistical significance. We offer recommendations for improving practice in empirical research based on these findings.

Tidying up a messy table may appear simple for humans, but articulating clear criteria for tidiness is challenging due to the ambiguous nature of common sense reasoning. Large Language Models (LLMs) have proven capable of capturing common sense knowledge to reason over this vague concept of tidiness. However, they alone may struggle with table tidying due to the limited grasp on the spatio-visual aspects of tidiness. In this work, we aim to ground the common-sense concept of tidiness within the context of object arrangement. Our survey reveals that humans usually factorize tidiness into semantic and visual-spatial tidiness; our grounding approach aligns with this decomposition. We connect a language-based policy generator with an image-based tidiness score function: the policy generator utilizes the LLM's commonsense knowledge to cluster objects by their implicit types and functionalities for semantic tidiness; meanwhile, the tidiness score function assesses the visual-spatial relations of the object to achieve visual-spatial tidiness. Our tidiness score is trained using synthetic data generated cheaply from customized random walks, which inherently encode the order of tidiness, thereby bypassing the need for labor-intensive human demonstrations. The simulated experiment shows that our approach successfully generates tidy arrangements, predominately in 2D, with potential for 3D stacking, for tables with various novel objects., Comment: RSSLRL2023 Workshop, Under review for conference

Inverse reinforcement learning (IRL) algorithms often rely on (forward) reinforcement learning or planning over a given time horizon to compute an approximately optimal policy for a hypothesized reward function and then match this policy with expert demonstrations. The time horizon plays a critical role in determining both the accuracy of reward estimate and the computational efficiency of IRL algorithms. Interestingly, an effective time horizon shorter than the ground-truth value often produces better results faster. This work formally analyzes this phenomenon and provides an explanation: the time horizon controls the complexity of an induced policy class and mitigates overfitting with limited data. This analysis leads to a principled choice of the effective horizon for IRL. It also prompts us to reexamine the classic IRL formulation: it is more natural to learn jointly the reward and the effective horizon together rather than the reward alone with a given horizon. Our experimental results confirm the theoretical analysis., Comment: 9 pages, under review

We experimentally demonstrate an all-optical random number generator based on spontaneous symmetry breaking in a coherently-driven Kerr resonator. Random bit sequences are generated by repeatedly tuning a control parameter across a symmetry-breaking bifurcation that enacts random selection between two possible steady-states of the system. Experiments are performed in a fibre ring resonator, where the two symmetry-broken steady-states are associated with orthogonal polarization modes. Detrimental biases due to system asymmetries are completely suppressed by leveraging a recently-discovered self-symmetrization phenomenon that ensures the symmetry breaking acts as an unbiased coin toss, with a genuinely random selection between the two available steady-states. We optically generate bits at a rate of over 3~MHz without post-processing and verify their randomness using the National Institute of Standards and Technology and Dieharder statistical test suites., Comment: 5 pages, 4 figures

Instrumental variable (IV) strategies are widely used in political science to establish causal relationships. However, the identifying assumptions required by an IV design are demanding, and it remains challenging for researchers to assess their validity. In this paper, we replicate 67 papers published in three top journals in political science during 2010-2022 and identify several troubling patterns. First, researchers often overestimate the strength of their IVs due to non-i.i.d. errors, such as a clustering structure. Second, the most commonly used t-test for the two-stage-least-squares (2SLS) estimates often severely underestimates uncertainty. Using more robust inferential methods, we find that around 19-30% of the 2SLS estimates in our sample are underpowered. Third, in the majority of the replicated studies, the 2SLS estimates are much larger than the ordinary-least-squares estimates, and their ratio is negatively correlated with the strength of the IVs in studies where the IVs are not experimentally generated, suggesting potential violations of unconfoundedness or the exclusion restriction. To help researchers avoid these pitfalls, we provide a checklist for better practice., Comment: Forthcoming in Political Analysis. Appendix (supp.pdf) in archived zip

Knowledge and skills can transfer from human teachers to human students. However, such direct transfer is often not scalable for physical tasks, as they require one-to-one interaction, and human teachers are not available in sufficient numbers. Machine learning enables robots to become experts and play the role of teachers to help in this situation. In this work, we formalize cooperative robot teaching as a Markov game, consisting of four key elements: the target task, the student model, the teacher model, and the interactive teaching-learning process. Under a moderate assumption, the Markov game reduces to a partially observable Markov decision process, with an efficient approximate solution. We illustrate our approach on two cooperative tasks, one in a simulated video game and one with a real robot., Comment: CoRL 2022

External driving of passive, nonlinear optical resonators has emerged over the past decade as a novel route for the generation of ultrashort optical pulses and corresponding broadband frequency combs. Whilst the pulse formation dynamics in such systems differ dramatically from those manifesting themselves in conventional mode-locked lasers, the demarcation between the two traditionally distinct paradigms has recently begun to blur, with demonstrations of hybrid systems incorporating both external driving and active media shown to offer specific advantages. Here we explore a new pathway for ultrashort pulse generation at the interface of externally-driven passive resonators and lasers. By leveraging the nonlinear Raman gain inherent to fused silica, we achieve deterministic generation of low-noise dissipative solitons with durations well below 100 fs via phase-coherent pulsed driving of resonators made of standard, commercially-available optical fibre. We explore and explain the physics of the new dissipative Raman soliton states, identifying scaling laws that govern the pulses' characteristics and that allow output repetition rates to be scaled at will without influencing the soliton duration. The scheme explored in our work enables the shortest ever pulses generated in resonators (active or passive) made from a single commercially-available optical fibre, and it has the potential to be transferred into a chip-scale format by using existing dispersion-engineered silica microresonators., Comment: 15 pages, 7 figures

Deformable Object Manipulation (DOM) is of significant importance to both daily and industrial applications. Recent successes in differentiable physics simulators allow learning algorithms to train a policy with analytic gradients through environment dynamics, which significantly facilitates the development of DOM algorithms. However, existing DOM benchmarks are either single-object-based or non-differentiable. This leaves the questions of 1) how a task-specific algorithm performs on other tasks and 2) how a differentiable-physics-based algorithm compares with the non-differentiable ones in general. In this work, we present DaXBench, a differentiable DOM benchmark with a wide object and task coverage. DaXBench includes 9 challenging high-fidelity simulated tasks, covering rope, cloth, and liquid manipulation with various difficulty levels. To better understand the performance of general algorithms on different DOM tasks, we conduct comprehensive experiments over representative DOM methods, ranging from planning to imitation learning and reinforcement learning. In addition, we provide careful empirical studies of existing decision-making algorithms based on differentiable physics, and discuss their limitations, as well as potential future directions., Comment: ICLR 2023 (Oral)

Inverse reinforcement learning (IRL) seeks to infer a cost function that explains the underlying goals and preferences of expert demonstrations. This paper presents receding horizon inverse reinforcement learning (RHIRL), a new IRL algorithm for high-dimensional, noisy, continuous systems with black-box dynamic models. RHIRL addresses two key challenges of IRL: scalability and robustness. To handle high-dimensional continuous systems, RHIRL matches the induced optimal trajectories with expert demonstrations locally in a receding horizon manner and 'stitches' together the local solutions to learn the cost; it thereby avoids the 'curse of dimensionality'. This contrasts sharply with earlier algorithms that match with expert demonstrations globally over the entire high-dimensional state space. To be robust against imperfect expert demonstrations and control noise, RHIRL learns a state-dependent cost function 'disentangled' from system dynamics under mild conditions. Experiments on benchmark tasks show that RHIRL outperforms several leading IRL algorithms in most instances. We also prove that the cumulative error of RHIRL grows linearly with the task duration., Comment: The paper is accepted by Conference on Neural Information Processing Systems (NeurIPS) 2022

Microcombs -- optical frequency combs generated in coherently-driven nonlinear microresonators -- have attracted significant attention over the last decade. The ability to generate two such combs in a single resonator device has in particular enabled a host of applications from spectroscopy to imaging. Concurrently, novel comb generation techniques such as synchronous pulsed driving have been developed to enhance the efficiency and flexibility of microcomb generation. Here we report on the first experimental demonstration of dual-microcomb generation via synchronous pulsed pumping of a single microresonator. Specifically, we use two electro-optically generated pulse trains derived from a common continuous wave laser to simultaneously drive two orthogonal polarization modes of an integrated silica ring resonator, observing the generation of coherent dissipative Kerr cavity soliton combs on both polarization axes. Thanks to the resonator birefringence, the two soliton combs are associated with different repetition rates, thus realizing a dual-microcomb source. To illustrate the source's application potential, we demonstrate proof-of-concept spectroscopic measurements.

Dissipative Kerr cavity solitons (CSs) are persisting pulses of light that manifest themselves in driven optical resonators and that have attracted significant attention over the last decade. Whilst the vast majority of studies have revolved around conditions where the resonator exhibits strong anomalous dispersion, recent studies have shown that solitons with unique characteristics and dynamics can arise under conditions of near-zero-dispersion driving. Here we report on experimental studies of the existence and stability dynamics of Kerr CSs under such conditions. In particular, we experimentally probe the solitons' range of existence and examine how their breathing instabilities are modified when group-velocity dispersion is close to zero, such that higher-order dispersion terms play a significant role. On the one hand, our experiments directly confirm earlier theoretical works that predict (i) breathing near-zero-dispersion solitons to emit polychromatic dispersive radiation, and (ii) that higher-order dispersion can extend the range over which the solitons are stable. On the other hand, our experiments also reveal a novel cross-over scenario, whereby the influence of higher-order dispersion changes from stabilising to destabilising. Our comprehensive experiments sample soliton dynamics both in the normal and anomalous dispersion regimes, and our results are in good agreement with numerical simulations and theoretical predictions., Comment: 8 pages, 6 figures

This paper introduces a simple framework of counterfactual estimation for causal inference with time-series cross-sectional data, in which we estimate the average treatment effect on the treated by directly imputing counterfactual outcomes for treated observations. We discuss several novel estimators under this framework, including the fixed effects counterfactual estimator, interactive fixed effects counterfactual estimator, and matrix completion estimator. They provide more reliable causal estimates than conventional twoway fixed effects models when treatment effects are heterogeneous or unobserved time-varying confounders exist. Moreover, we propose a new dynamic treatment effects plot, along with several diagnostic tests, to help researchers gauge the validity of the identifying assumptions. We illustrate these methods with two political economy examples and develop an open-source package, fect, in both R and Stata to facilitate implementation.

Designing a network to learn a molecule structure given its physical/chemical properties is a hard problem, but is useful for drug discovery tasks. In this paper, we incorporate higher-order relational learning of Factor Graphs with strong approximation power of Neural Networks to create a molecule-structure learning network that has strong generalization power and can enforce higher-order relationship and valence constraints. We further propose methods to tackle problems such as the efficient design of factor nodes, conditional parameter sharing among factors, and symmetry problems in molecule structure prediction. Our experiment evaluation shows that the factor learning is effective and outperforms related methods.

Kerr microresonators driven in the normal dispersion regime typically require the presence of localized dispersion perturbations, such as those induced by avoided mode crossings, to initiate the formation of optical frequency combs. In this work, we experimentally demonstrate that this requirement can be lifted by driving the resonator with a pulsed pump source. We also show that controlling the desynchronization between the pump repetition rate and the cavity free-spectral range (FSR) provides a simple mechanism to tune the center frequency of the output comb. Using a fiber mini-resonator with a radius of only 6 cm we experimentally present spectrally flat combs with a bandwidth of 3 THz whose center frequency can be tuned by more than 2 THz. By driving the cavity at harmonics of its 0.54 GHz FSR, we are able to generate combs with line spacings selectable between 0.54 and 10.8 GHz. The ability to tune both the center frequency and frequency spacing of the output comb highlights the flexibility of this platform. Additionally, we demonstrate that under conditions of large pump-cavity desynchronization, the same cavity also supports a new form of Raman-assisted anomalous dispersion cavity soliton.

With demonstrated applications ranging from metrology to telecommunications, soliton microresonator frequency combs have emerged over the past decade as a remarkable new technology. However, standard implementations only allow for the generation of combs whose repetition rate is tied close to the fundamental resonator free-spectral range (FSR), offering little or no dynamic control over the comb line spacing. Here we propose and experimentally demonstrate harmonic and rational harmonic driving as novel techniques that allow for the robust generation of soliton frequency combs with discretely adjustable frequency spacing. By driving an integrated Kerr microresonator with a periodic train of picosecond pulses whose repetition rate is set close to an integer harmonic of the 3.23 GHz cavity FSR, we deterministically generate soliton frequency combs with frequency spacings discretely adjustable between 3.23 GHz and 19.38 GHz. More remarkably, we also demonstrate that driving the resonator at rational fractions of the FSR allows for the generation of combs whose frequency spacing corresponds to an integer harmonic of the pump repetition rate. By measuring the combs' radio-frequency spectrum, we confirm operation in the low-noise soliton regime with no supermode noise. The novel techniques demonstrated in our work provide new degrees of freedom for the design of synchronously pumped soliton frequency combs., Comment: 7 pages, 5 figures

We report on the experimental observation of internally-pumped parametric oscillation in a high-Q lithium niobate microresonator under conditions of natural phase-matching. Specifically, launching near-infrared pump light around 1060 nm into a $z$-cut congruent lithium niobate microresonator, we observe the generation of optical sidebands around the input pump under conditions where second-harmonic generation is close to natural phase-matching. We find that a wide range of different sideband frequency shifts can be generated by varying the experimental parameters. Under particular conditions, we observe the cascaded generation of several equally-spaced sidebands around the pump -- the first steps of optical frequency comb generation via cavity-enhanced second-harmonic generation., Comment: 5 pages, 5 figures; a wider range of indicative results added to Fig. 3 and text changed accordingly in version 2

Chimera states -- named after the mythical beast with a lion's head, a goat's body, and a dragon's tail -- correspond to spatiotemporal patterns characterised by the coexistence of coherent and incoherent domains in coupled systems. They were first identified in 2002 in theoretical studies of spatially extended networks of Stuart-Landau oscillators, and have been subject to extensive theoretical and experimental research ever since. While initially considered peculiar to networks with weak nonlocal coupling, recent theoretical studies have predicted that chimera-like states can emerge even in systems with purely local coupling. Here we report on the first experimental observations of chimera-like states in a system with local coupling -- a coherently-driven Kerr nonlinear optical resonator. We show that artificially engineered discreteness -- realised by suitably modulating the coherent driving field -- allows for the nonlinear localisation of spatiotemporal complexity, and we demonstrate unprecedented control over the existence, characteristics, and dynamics of the resulting chimera-like states. Moreover, using ultrafast time lens imaging, we resolve the chimeras' picosecond-scale internal structure in real time., Comment: 9 pages, 5 figures

Frequency to time mapping is a powerful technique for observing ultrafast phenomena and non-repetitive events in optics. However, many optical sources operate in wavelength regions, or at power levels, that are not compatible with standard frequency to time mapping implementations. The recently developed free-space angular chirp enhanced delay (FACED) removes many of these limitations, and offers a linear frequency to time mapping in any wavelength region where high-reflectivity mirrors and diffractive optics are available. In this work, we present a detailed formulation of the optical transfer function of a FACED device. Experimentally, we verify the properties of this transfer function, and then present simple guidelines to guarantee the correct operation of a FACED frequency to time measurement. We also experimentally demonstrate the real-time spectral analysis of femtosecond and picosecond pulses using this system.