Search

The "LLM-as-a-judge" paradigm employs Large Language Models (LLMs) as annotators and evaluators in tasks traditionally performed by humans. LLM annotations are widely used, not only in NLP research but also in fields like medicine, psychology, and social science. Despite their role in shaping study results and insights, there is no standard or rigorous procedure to determine whether LLMs can replace human annotators. In this paper, we propose a novel statistical procedure -- the Alternative Annotator Test (alt-test) -- that requires only a modest subset of annotated examples to justify using LLM annotations. Additionally, we introduce a versatile and interpretable measure for comparing LLM judges. To demonstrate our procedure, we curated a diverse collection of ten datasets, consisting of language and vision-language tasks, and conducted experiments with six LLMs and four prompting techniques. Our results show that LLMs can sometimes replace humans with closed-source LLMs (such as GPT-4o), outperforming open-source LLMs, and that prompting techniques yield judges of varying quality. We hope this study encourages more rigorous and reliable practices.

Scintillation, the process of converting high-energy radiation to detectable visible light, is pivotal in advanced technologies spanning from medical diagnostics to fundamental scientific research. Despite significant advancements toward faster and more efficient scintillators, there remains a fundamental limit arising from the intrinsic properties of scintillating materials. The scintillation process culminates in spontaneous emission of visible light, which is restricted in rate by the oscillator strength of individual emission centers. Here, we observe a novel collective emission phenomenon under X-ray excitation, breaking this limit and accelerating the emission. Our observation reveals that strong interactions between simultaneously excited coupled perovskite quantum dots can create collective radioluminescence. This effect is characterized by a spectral shift and an enhanced rate of emission, with an average lifetime of 230 ps, 14 times faster than their room temperature spontaneous emission. It has been established that such quantum dots exhibit superfluorescence under UV excitation. However, X-ray superfluorescence is inherently different, as each high-energy photon creates multiple synchronized excitation events, triggered by a photoelectron and resulting in even faster emission rates, a larger spectral shift, and a broader spectrum. This observation is consistent with a quantum-optical analysis explaining both the UV-driven and X-ray-driven effects. We use a Hanbury-Brown-Twiss g^(2) ({\tau}) setup to analyze the temperature-dependent temporal response of these scintillators. Collective radioluminescence breaks the limit of scintillation lifetime based on spontaneous emission and could dramatically improve time-of-flight detector performance, introducing quantum enhancements to scintillation science.

Hanbury Brown and Twiss (HBT) interferometry is a milestone experiment that transformed our understanding of the nature of light. The concept was demonstrated in 1956 to measure the radii of stars through photon coincidence detection. This form of coincidence detection later became a cornerstone of modern quantum optics. Here we connect HBT interferometry to the physics of scintillation, the process of spontaneous light emission upon excitation by high-energy particles, such as x-rays. Our work reveals intrinsic photon bunching in the scintillation process, which we utilize to elucidate its underlying light emission mechanisms. Specifically, g^((2) ) ({\tau}) enables the quantitative extraction of scintillation lifetime and light yield, showing their dependence on temperature and X-ray flux as well. This approach provides a characterization method that we benchmark on a wide gamut of scintillators, including rare-earth-doped garnets and perovskite nanocrystals. Our method is particularly important for nano- and micro-scale scintillators, whose properties are challenging to quantify by conventional means: We extract the scintillation properties in perovskite nanocrystals of only a few hundreds of nanometers, observing strong photon bunching (g^((2) ) (0)>50). Our research paves the way for broader use of photon-coincidence measurement and methods from quantum optics in studying materials with complex optical properties in extremes regions of the electromagnetic spectrum.

Ongoing efforts that span over decades show a rise of AI methods for accelerating scientific discovery, yet accelerating discovery in mathematics remains a persistent challenge for AI. Specifically, AI methods were not effective in creation of formulas for mathematical constants because each such formula must be correct for infinite digits of precision, with "near-true" formulas providing no insight toward the correct ones. Consequently, formula discovery lacks a clear distance metric needed to guide automated discovery in this realm. In this work, we propose a systematic methodology for categorization, characterization, and pattern identification of such formulas. The key to our methodology is introducing metrics based on the convergence dynamics of the formulas, rather than on the numerical value of the formula. These metrics enable the first automated clustering of mathematical formulas. We demonstrate this methodology on Polynomial Continued Fraction formulas, which are ubiquitous in their intrinsic connections to mathematical constants, and generalize many mathematical functions and structures. We test our methodology on a set of 1,768,900 such formulas, identifying many known formulas for mathematical constants, and discover previously unknown formulas for $\pi$, $\ln(2)$, Gauss', and Lemniscate's constants. The uncovered patterns enable a direct generalization of individual formulas to infinite families, unveiling rich mathematical structures. This success paves the way towards a generative model that creates formulas fulfilling specified mathematical properties, accelerating the rate of discovery of useful formulas., Comment: 8 figures, 5 tables, 28 pages including the supplementary information. For a 5-minute video abstract see https://recorder-v3.slideslive.com/#/share?share=97010&s=c47967e3-d585-453c-a4dd-a4fa7955dba3 . Code can be found at https://github.com/RamanujanMachine/Blind-Delta-Algorithm

Directional couplers are essential components in integrated photonics. Given their widespread use, accurate characterization of directional couplers is crucial for ensuring optimal performance. However, it is challenging due to the coupling between fibers and waveguides, which is highly sensitive to alignment and fabrication imperfections. To address these challenges, we propose a novel direct measurement technique that offers greater robustness to variations in optical interfaces, while bypassing extinction ratio measurements. Our method enables a broadband and precise characterization of the directional couplers' splitting ratio. We experimentally validate this approach, demonstrate its robustness against intentional errors, and compare it to a naive direct measurement method. Furthermore, our technique is generalized to measure the amplitude of any general 2x2 unitary circuit, providing valuable insights for designing and testing a wide range of photonic integrated devices.

In this paper we study a quantum version of the multiparty simultaneous message-passing (SMP) model, and we show that in some cases, quantum communication can replace public randomness, even with no entanglement between the parties. This was already known for two players, but not for more than two players, and indeed, so far all that was known was a negative result. Our main technical contribution is a compiler that takes any classical public-coin simultaneous protocol based on "modified equality queries," and converts it into a quantum simultaneous protocol without public coins with roughly the same communication complexity. We then use our compiler to derive protocols for several problems, including frequency moments, neighborhood diversity, enumeration of isolated cliques, and more., Comment: 20 pages

We present a fully calculable UV complete toy model of a Peccei-Quinn (PQ) axion coupled to magnetic monopoles as well as electric charges. The theory has manifest electric-magnetic duality built in. We find that the axion-photon coupling contains the usual anomaly term, plus periodic corrections which can also become large if the monopole is light, without violating the discrete axion shift symmetry. These additional periodic terms can be identified as the non-perturbative corrections due to the monopoles (and other BPS states), but can also be interpreted as a sum over instanton corrections. The key aspect helping reconcile axion coupling quantization with electric-magnetic duality is the fact that the axion itself undergoes a non-linear transformation under electric-magnetic duality. The theory analyzed here is just the original $N=2$ supersymmetric $SU(2)$ Seiberg-Witten theory, which contains a PQ axion due to an anomalous spontaneously broken global $R$-symmetry, as well as massless fermionic monopoles and dyons at special points in the moduli space. Hence the entire machinery of the Seiberg-Witten solution can be applied to reliably calculate the photon-axion coupling in different duality frames. We show explicitly that the physically observable axion-photon amplitude is duality invariant, as it had to be., Comment: 10 pages, 5 figures

When artificial agents are jointly trained to perform collaborative tasks using a communication channel, they develop opaque goal-oriented communication protocols. Good task performance is often considered sufficient evidence that meaningful communication is taking place, but existing empirical results show that communication strategies induced by common objectives can be counterintuitive whilst solving the task nearly perfectly. In this work, we identify a goal-agnostic prerequisite to meaningful communication, which we term semantic consistency, based on the idea that messages should have similar meanings across instances. We provide a formal definition for this idea, and use it to compare the two most common objectives in the field of emergent communication: discrimination and reconstruction. We prove, under mild assumptions, that semantically inconsistent communication protocols can be optimal solutions to the discrimination task, but not to reconstruction. We further show that the reconstruction objective encourages a stricter property, spatial meaningfulness, which also accounts for the distance between messages. Experiments with emergent communication games validate our theoretical results. These findings demonstrate an inherent advantage of distance-based communication goals, and contextualize previous empirical discoveries., Comment: 34 pages, to be published in NeurIPS 2024

Accurate human pose estimation is essential for effective Human-Robot Interaction (HRI). By observing a user's arm movements, robots can respond appropriately, whether it's providing assistance or avoiding collisions. While visual perception offers potential for human pose estimation, it can be hindered by factors like poor lighting or occlusions. Additionally, wearable inertial sensors, though useful, require frequent calibration as they do not provide absolute position information. Force-myography (FMG) is an alternative approach where muscle perturbations are externally measured. It has been used to observe finger movements, but its application to full arm state estimation is unexplored. In this letter, we investigate the use of a wearable FMG device that can observe the state of the human arm for real-time applications of HRI. We propose a Transformer-based model to map FMG measurements from the shoulder of the user to the physical pose of the arm. The model is also shown to be transferable to other users with limited decline in accuracy. Through real-world experiments with a robotic arm, we demonstrate collision avoidance without relying on visual perception.

Genome analysis has revolutionized fields such as personalized medicine and forensics. Modern sequencing machines generate vast amounts of fragmented strings of genome data called reads. The alignment of these reads into a complete DNA sequence of an organism (the read mapping process) requires extensive data transfer between processing units and memory, leading to execution bottlenecks. Prior studies have primarily focused on accelerating specific stages of the read-mapping task. Conversely, this paper introduces a holistic framework called DART-PIM that accelerates the entire read-mapping process. DART-PIM facilitates digital processing-in-memory (PIM) for an end-to-end acceleration of the entire read-mapping process, from indexing using a unique data organization schema to filtering and read alignment with an optimized Wagner Fischer algorithm. A comprehensive performance evaluation with real genomic data shows that DART-PIM achieves a 5.7x and 257x improvement in throughput and a 92x and 27x energy efficiency enhancement compared to state-of-the-art GPU and PIM implementations, respectively.

Ensuring food security is a critical global challenge, particularly for low-income countries where food prices impact the access to nutritious food. The volatility of global agricultural commodity (AC) prices exacerbates food insecurity, with international trade restrictions and market disruptions further complicating the situation. Despite online platforms for monitoring food prices, there is a significant gap in providing detailed explanations and forecasts accessible to non-specialists. To address this, we propose the Agricultural Commodity Analysis and Forecasts (AGRICAF) methodology, integrating explainable machine learning (XML) and econometric techniques to analyse and forecast global agricultural commodity prices up to one year ahead, dynamically adapting to different forecast horizons. This innovative integration allows us to model complex interactions and dynamics while providing clear, interpretable results. This paper demonstrates how AGRICAF can be used, applying it to three major agricultural commodities - maize, soybean, and wheat - and explaining how different factors impact prices across various months and forecast horizons. By facilitating access to accurate and interpretable medium-term forecasts of AC prices, AGRICAF can contribute to developing a fair and sustainable food system., Comment: Comments, suggestions, criticism or questions are more than welcome! Feel free to email me

The lack of automatic evaluation metrics tailored for SignWriting presents a significant obstacle in developing effective transcription and translation models for signed languages. This paper introduces a comprehensive suite of evaluation metrics specifically designed for SignWriting, including adaptations of standard metrics such as \texttt{BLEU} and \texttt{chrF}, the application of \texttt{CLIPScore} to SignWriting images, and a novel symbol distance metric unique to our approach. We address the distinct challenges of evaluating single signs versus continuous signing and provide qualitative demonstrations of metric efficacy through score distribution analyses and nearest-neighbor searches within the SignBank corpus. Our findings reveal the strengths and limitations of each metric, offering valuable insights for future advancements using SignWriting. This work contributes essential tools for evaluating SignWriting models, facilitating progress in the field of sign language processing. Our code is available at \url{https://github.com/sign-language-processing/signwriting-evaluation}.

We formulate a complex analog of the celebrated Levi-Hadwiger-Boltyanski illumination (or covering) conjecture for complex convex bodies in C^n, as well as its (non-comparable) fractional version. A key element in posing these problems is computing the classical and fractional illumination numbers of the complex analog of the hypercube, i.e., the polydisc. We prove that the illumination number of the polydisc in C^n is equal to 2^(n+1)-1 and that the fractional illumination number of the polydisc in C^n is 2^n. In addition, we verify both conjectures for the classes of complex zonotopes and zonoids., Comment: 35 pages, 4 figures

The open-ended coaxial probe (OECP) technique is one of the most commonly used methods for the characterization of homogeneous media properties, especially in the biomedical sciences. However, when considering inhomogeneous media, the effect of the heterogeneity on the probe terminal admittance is unclear, making the measured admittance hard to interpret and relate to the medium properties. In this paper we present an analytical model for the contribution of an isotropic scatterer embedded in an otherwise homogeneous medium to the probe admittance. We utilize rigorous scattering theory and various approximations to obtain simplified, closed-form expressions. Using the obtained results we present a method to accurately extract the scatterer properties from a measurement of the admittance. In addition, we define the sensing depth, and show how it can be mapped as a function of the expected scatterer properties. Full-wave simulations are used to verify the analytical model, and the proposed method paves a path for further generalization to additional scenarios of open-coaxial probe sensing of an inhomogeneous medium., Comment: 7 pages, 6 figures

In this paper, we study new extensions of the functional Blaschke-Santalo inequalities, and explore applications of such new inequalities beyond the classical setting of the standard Gaussian measure.

We find a family of groups generated by a pair of parabolic elements in which every relator must admit a long subword of a specific form. In particular, this collection contains groups in which the number of syllables of any relator is arbitrarily large. This suggests that the existing methods for finding non-free groups with rational parabolic generators may be inadequate in this case, as they depend on the presence of relators with few syllables. Our results rely on two variants of the ping-pong lemma that we develop, applicable to groups that are possibly non-free. These variants aim to isolate the group elements responsible for the failure of the classical ping-pong lemma., Comment: 16 pages, 2 figures

We study the Brunn-Minkowski and Borell-Brascamp-Lieb inequalities on a Riemannian manifold endowed with an exact magnetic field, replacing geodesics by minimizers of an action functional given by the length minus the integral of the magnetic potential. We prove that nonnegativity of a suitably defined magnetic Ricci curvature is equivalent to a magnetic Brunn-Minkowski inequality. More generally, given an arbitrary volume form on the manifold, we introduce a weighted magnetic Ricci curvature, and prove a magnetic version of the Borell-Brascamp-Lieb inequality., Comment: 55 pages, 1 figure. A converse to the main theorem was added. Higher dimensional complex hyperbolic space was added as an example

Magnetic resonance imaging (MRI) is crucial in diagnosing various abdominal conditions and anomalies. Traditional MRI scans often yield anisotropic data due to technical constraints, resulting in varying resolutions across spatial dimensions, which limits diagnostic accuracy and volumetric analysis. Super-resolution (SR) techniques aim to address these limitations by reconstructing isotropic high-resolution images from anisotropic data. However, current SR methods often rely on indirect mappings and limited training data, focusing mainly on two-dimensional improvements rather than achieving true three-dimensional isotropy. We introduce SIMPLE, a Simultaneous Multi-Plane Self-Supervised Learning approach for isotropic MRI restoration from anisotropic data. Our method leverages existing anisotropic clinical data acquired in different planes, bypassing the need for simulated downsampling processes. By considering the inherent three-dimensional nature of MRI data, SIMPLE ensures realistic isotropic data generation rather than solely improving through-plane slices. This approach flexibility allows it to be extended to multiple contrast types and acquisition methods commonly used in clinical settings. Our experiments show that SIMPLE outperforms state-of-the-art methods both quantitatively using the Kernel Inception Distance (KID) and semi-quantitatively through radiologist evaluations. The generated isotropic volume facilitates more accurate volumetric analysis and 3D reconstructions, promising significant improvements in clinical diagnostic capabilities.

By 2050, the global population will exceed 9 billion, demanding a 70% increase in food production. Animal proteins alone may not suffice and contribute to global warming. Alternative proteins such as legumes, algae, and insects are being explored, but their health impacts are largely unknown. For this, three-week-old rats were fed diets containing 20% protein from various sources for six weeks. A casein-based control diet was compared to soy isolate, spirulina powder, chickpea isolate, chickpea flour, and fly larvae powder. Except for spirulina, alternative protein groups showed comparable growth patterns to the casein group. Morphological and mechanical tests of femur bones matched growth patterns. Caecal 16S analysis highlighted the impact on gut microbiota diversity. Chickpea flour showed significantly lower $\alpha$-diversity compared with casein and chickpea isolate groups while chickpea flour, had the greatest distinction in $\beta$-diversity. Alternative protein sources supported optimal growth, but quality and health implications require further exploration.

We prove two special cases of a strengthened Gaussian correlation conjecture, due to Tehranchi, and show that if the conjecture holds asymptotically, it holds for any dimension. Additionally, we use these special cases to prove a refined version of the \v{S}id\'ak-Khatri inequality.

In this paper we prove that for stable semi-discretizations of the wave equation for the WaveHoltz iteration is guaranteed to converge to an approximate solution of the corresponding frequency domain problem, if it exists. We show that for certain classes of frequency domain problems, the WaveHoltz iteration without acceleration converges in $O({\omega})$ iterations with the constant factor depending logarithmically on the desired tolerance. We conjecture that the Helmholtz problem in open domains with no trapping waves is one such class of problems and we provide numerical examples in one and two dimensions using finite differences and discontinuous Galerkin discretizations which demonstrate these converge results.

Continuous variable quantum key distribution (CVQKD) with discrete modulation combines advantages of CVQKD, such as the implementability using readily available technologies, with advantages of discrete variable quantum key distribution, such as easier error correction procedures. We consider a prepare-and-measure CVQKD protocol, where Alice chooses from a set of four coherent states and Bob performs a heterodyne measurement, the result of which is discretised in both key and test rounds. We provide a security proof against coherent attacks in the finite-size regime, and compute the achievable key rate. To this end, we employ the generalised entropy accumulation theorem, as well as recent advances in conic optimisation, yielding improved key rates compared to previous works. At metropolitan distances, our method can provide positive key rates for the order of $10^8$ rounds., Comment: 24 pages, 5 figures

Forced alignment (FA) plays a key role in speech research through the automatic time alignment of speech signals with corresponding text transcriptions. Despite the move towards end-to-end architectures for speech technology, FA is still dominantly achieved through a classic GMM-HMM acoustic model. This work directly compares alignment performance from leading automatic speech recognition (ASR) methods, WhisperX and Massively Multilingual Speech Recognition (MMS), against a Kaldi-based GMM-HMM system, the Montreal Forced Aligner (MFA). Performance was assessed on the manually aligned TIMIT and Buckeye datasets, with comparisons conducted only on words correctly recognized by WhisperX and MMS. The MFA outperformed both WhisperX and MMS, revealing a shortcoming of modern ASR systems. These findings highlight the need for advancements in forced alignment and emphasize the importance of integrating traditional expertise with modern innovation to foster progress. Index Terms: forced alignment, phoneme alignment, word alignment

We recently presented DISentangled COunterfactual Visual interpretER (DISCOVER), a method toward systematic visual interpretability of image-based classification models and demonstrated its applicability to two biomedical domains. Here we demonstrate that DISCOVER can be applied to the domain of natural images. First, DISCOVER visually interpreted the nose size, the muzzle area, and the face size as semantic discriminative visual traits discriminating between facial images of dogs versus cats. Second, DISCOVER visually interpreted the cheeks and jawline, eyebrows and hair, and the eyes, as discriminative facial characteristics. These successful visual interpretations across two natural images domains indicate that DISCOVER is a generalized interpretability method.

Diffusion-based generative models have recently shown remarkable image and video editing capabilities. However, local video editing, particularly removal of small attributes like glasses, remains a challenge. Existing methods either alter the videos excessively, generate unrealistic artifacts, or fail to perform the requested edit consistently throughout the video. In this work, we focus on consistent and identity-preserving removal of glasses in videos, using it as a case study for consistent local attribute removal in videos. Due to the lack of paired data, we adopt a weakly supervised approach and generate synthetic imperfect data, using an adjusted pretrained diffusion model. We show that despite data imperfection, by learning from our generated data and leveraging the prior of pretrained diffusion models, our model is able to perform the desired edit consistently while preserving the original video content. Furthermore, we exemplify the generalization ability of our method to other local video editing tasks by applying it successfully to facial sticker-removal. Our approach demonstrates significant improvement over existing methods, showcasing the potential of leveraging synthetic data and strong video priors for local video editing tasks.

It is usually assumed that $4D$ instantons can only arise in non-Abelian theories. In this paper we re-examine this conventional wisdom by explicitly constructing instantons in an Abelian gauge theory: ${\rm QED}_4$ with $N_f$ flavors of Dirac fermions, in the background of a Dirac monopole. This is the low-energy effective field theory for fermions interacting with a 't Hooft-Polyakov monopole, in the limit where the monopole is infinitely heavy (hence pointlike) and static. This theory, whose non-topological sectors were studied by Rubakov and Callan, has a far richer structure than previously explored. We show how to calculate the topological instanton number, demonstrate the existence of 't Hooft zero modes localized around such instantons, and show how instantons in the path integral provide the underlying mechanism for the Callan-Rubakov process: monopole-catalyzed baryon decay with a cross section that saturates the unitarity bound. Our computation relies on correctly identifying the relevant $2D$ EFT for monopole catalysis as Axial ${\rm QED}_2$ in an effective $AdS_2$ metric., Comment: 43 pages main text, 12 pages appendix, 4 figures

Given the remarkable results of motion synthesis with diffusion models, a natural question arises: how can we effectively leverage these models for motion editing? Existing diffusion-based motion editing methods overlook the profound potential of the prior embedded within the weights of pre-trained models, which enables manipulating the latent feature space; hence, they primarily center on handling the motion space. In this work, we explore the attention mechanism of pre-trained motion diffusion models. We uncover the roles and interactions of attention elements in capturing and representing intricate human motion patterns, and carefully integrate these elements to transfer a leader motion to a follower one while maintaining the nuanced characteristics of the follower, resulting in zero-shot motion transfer. Editing features associated with selected motions allows us to confront a challenge observed in prior motion diffusion approaches, which use general directives (e.g., text, music) for editing, ultimately failing to convey subtle nuances effectively. Our work is inspired by how a monkey closely imitates what it sees while maintaining its unique motion patterns; hence we call it Monkey See, Monkey Do, and dub it MoMo. Employing our technique enables accomplishing tasks such as synthesizing out-of-distribution motions, style transfer, and spatial editing. Furthermore, diffusion inversion is seldom employed for motions; as a result, editing efforts focus on generated motions, limiting the editability of real ones. MoMo harnesses motion inversion, extending its application to both real and generated motions. Experimental results show the advantage of our approach over the current art. In particular, unlike methods tailored for specific applications through training, our approach is applied at inference time, requiring no training. Our webpage is at https://monkeyseedocg.github.io., Comment: Video: https://www.youtube.com/watch?v=s5oo3sKV0YU, Project page: https://monkeyseedocg.github.io, Code: https://github.com/MonkeySeeDoCG/MoMo-code

Recent developments in static patch holography proposed that quantum gravity in de Sitter space admits a dual description in terms of a quantum mechanical theory living on a timelike surface near the cosmological horizon. In parallel, geometric observables associated with the Einstein-Rosen bridge of a black hole background were suggested to compute the computational complexity of the state dual to a gravitational theory. In this work, we pursue the study of the complexity=volume and complexity=action conjectures in a Schwarzschild-de Sitter geometry perturbed by the insertion of a shockwave at finite boundary times. This analysis extends previous studies that focused either on the complexity=volume 2.0 conjecture, or on the case of a shockwave inserted along the cosmological horizon. We show that the switchback effect, describing the delay in the evolution of complexity in reaction to a perturbation, is a universal feature of the complexity proposals in asymptotically de Sitter space. The geometric origin of this phenomenon is related to the causal connection between the static patches of de Sitter space when a positive pulse of null energy is inserted in the geometry., Comment: 42 pages+ appendices, 22 figures; v2: references added, minor changes

When solving ill-posed inverse problems, one often desires to explore the space of potential solutions rather than be presented with a single plausible reconstruction. Valuable insights into these feasible solutions and their associated probabilities are embedded in the posterior distribution. However, when confronted with data of high dimensionality (such as images), visualizing this distribution becomes a formidable challenge, necessitating the application of effective summarization techniques before user examination. In this work, we introduce a new approach for visualizing posteriors across multiple levels of granularity using tree-valued predictions. Our method predicts a tree-valued hierarchical summarization of the posterior distribution for any input measurement, in a single forward pass of a neural network. We showcase the efficacy of our approach across diverse datasets and image restoration challenges, highlighting its prowess in uncertainty quantification and visualization. Our findings reveal that our method performs comparably to a baseline that hierarchically clusters samples from a diffusion-based posterior sampler, yet achieves this with orders of magnitude greater speed., Comment: 32 pages, 21 figures

A data structure is called history independent if its internal memory representation does not reveal the history of operations applied to it, only its current state. In this paper we study history independence for concurrent data structures, and establish foundational possibility and impossibility results. We show that a large class of concurrent objects cannot be implemented from smaller base objects in a manner that is both wait-free and history independent; but if we settle for either lock-freedom instead of wait-freedom or for a weak notion of history independence, then at least one object in the class, multi-valued single-reader single-writer registers, can be implemented from smaller base objects, binary registers. On the other hand, using large base objects, we give a strong possibility result in the form of a universal construction: an object with $s$ possible states can be implemented in a wait-free, history-independent manner from compare-and-swap base objects that each have $O(s + 2^n)$ possible memory states, where $n$ is the number of processes in the system.

Langevin dynamics (LD) is widely used for sampling from distributions and for optimization. In this work, we derive a closed-form expression for the expected loss of preconditioned LD near stationary points of the objective function. We use the fact that at the vicinity of such points, LD reduces to an Ornstein-Uhlenbeck process, which is amenable to convenient mathematical treatment. Our analysis reveals that when the preconditioning matrix satisfies a particular relation with respect to the noise covariance, LD's expected loss becomes proportional to the rank of the objective's Hessian. We illustrate the applicability of this result in the context of neural networks, where the Hessian rank has been shown to capture the complexity of the predictor function but is usually computationally hard to probe. Finally, we use our analysis to compare SGD-like and Adam-like preconditioners and identify the regimes under which each of them leads to a lower expected loss., Comment: Accepted to AAAI-24 main track

TestGen automatically generates unit tests, carved from serialized observations of complex objects, observed during app execution. We describe the development and deployment of TestGen at Meta. In particular, we focus on the scalability challenges overcome during development in order to deploy observation-based test carving at scale in industry. So far, TestGen has landed 518 tests into production, which have been executed 9,617,349 times in continuous integration, finding 5,702 faults. Meta is currently in the process of more widespread deployment. Our evaluation reveals that, when carving its observations from 4,361 reliable end-to-end tests, TestGen was able to generate tests for at least 86\% of the classes covered by end-to-end tests. Testing on 16 Kotlin Instagram app-launch-blocking tasks demonstrated that the TestGen tests would have trapped 13 of these before they became launch blocking., Comment: 12 pages, 8 figures, FSE 2024, Mon 15 - Fri 19 July 2024, Porto de Galinhas, Brazil