Search

The Poltyrev bound provides a very tight upper bound on the decoding error probability when using binary linear codes for transmission over the binary symmetric channel and the additive white Gaussian noise channel, making use of the code's weight spectrum. In the present work, the bound is extended to memoryless symmetric channels with a discrete output alphabet. The derived bound is demonstrated on a hybrid BSC-BEC channel. Additionally, a reduced-complexity bound is introduced at the cost of some loss in tightness., Comment: 6 pages,8 figures

We explore the evolving relationship between galaxies and their dark matter halos from $z \sim 0.1$ to $z \sim 12$ using mass-limited angular clustering measurements in the 0.54 deg$^2$ of the COSMOS-Web survey. This study provides the first measurements of the mass-limited two-point correlation function at $z \ge 10$ and a consistent analysis spanning 13.4 Gyr of cosmic history, setting new benchmarks for future simulations and models. Using a halo occupation distribution (HOD) framework, we derive characteristic halo masses and the stellar-to-halo mass relationship (SHMR) across redshifts and stellar mass bins. Our results first indicate that HOD models fit data at $z \ge 2.5$ best when incorporating a non-linear scale-dependent halo bias, boosting clustering at non-linear scales (r = 10-100 kpc). We find that galaxies at z > 10.5 with $\log(M_\star / M_\odot) \ge 8.85$ are hosted by halos with $M_{\rm h} \sim 10^{10.5}\,M_\odot$, achieving a star formation efficiency (SFE) $M_\star / (f_b M_{\rm h}) $ up to 1 dex higher than at $z \le 1$. The high galaxy bias at $z \ge 8$ suggests that these galaxies reside in massive halos with intrinsic high SFE. Our SHMR evolves significantly with redshift, starting high at $z \ge 10.5$, decreasing until $z \sim 2 - 3$, then increasing again until the present. Current simulations fail to reproduce both massive high-$z$ galaxies and this evolution, while semi-empirical models linking SFE to halo mass, accretion rates, and redshift align with our findings. We propose that $z > 8$ galaxies experience bursty star formation without significant feedback altering their growth, driving the rapid growth of massive galaxies observed by JWST. Over time, increasing feedback efficiency and exponential halo growth suppress star formation. At $z \sim 2 - 3$ and after, halo growth slows down while star formation continues, supported by gas reservoirs in halos., Comment: Submitted to Astronomy & Astrophysics

Reliably generating structured outputs has become a critical capability for modern language model (LM) applications. Constrained decoding has emerged as the dominant technology across sectors for enforcing structured outputs during generation. Despite its growing adoption, little has been done with the systematic evaluation of the behaviors and performance of constrained decoding. Constrained decoding frameworks have standardized around JSON Schema as a structured data format, with most uses guaranteeing constraint compliance given a schema. However, there is poor understanding of the effectiveness of the methods in practice. We present an evaluation framework to assess constrained decoding approaches across three critical dimensions: efficiency in generating constraint-compliant outputs, coverage of diverse constraint types, and quality of the generated outputs. To facilitate this evaluation, we introduce JSONSchemaBench, a benchmark for constrained decoding comprising 10K real-world JSON schemas that encompass a wide range of constraints with varying complexity. We pair the benchmark with the existing official JSON Schema Test Suite and evaluate six state-of-the-art constrained decoding frameworks, including Guidance, Outlines, Llamacpp, XGrammar, OpenAI, and Gemini. Through extensive experiments, we gain insights into the capabilities and limitations of constrained decoding on structured generation with real-world JSON schemas. Our work provides actionable insights for improving constrained decoding frameworks and structured generation tasks, setting a new standard for evaluating constrained decoding and structured generation. We release JSONSchemaBench at https://github.com/guidance-ai/jsonschemabench

Deep neural networks suffer from storing millions and billions of weights in memory post-training, making challenging memory-intensive models to deploy on embedded devices. The weight-sharing technique is one of the popular compression approaches that use fewer weight values and share across specific connections in the network. In this paper, we propose a multi-objective evolutionary algorithm (MOEA) based compression framework independent of neural network architecture, dimension, task, and dataset. We use uniformly sized bins to quantize network weights into a single codebook (lookup table) for efficient weight representation. Using MOEA, we search for Pareto optimal $k$ bins by optimizing two objectives. Then, we apply the iterative merge technique to non-dominated Pareto frontier solutions by combining neighboring bins without degrading performance to decrease the number of bins and increase the compression ratio. Our approach is model- and layer-independent, meaning the weights are mixed in the clusters from any layer, and the uniform quantization method used in this work has $O(N)$ complexity instead of non-uniform quantization methods such as k-means with $O(Nkt)$ complexity. In addition, we use the center of clusters as the shared weight values instead of retraining shared weights, which is computationally expensive. The advantage of using evolutionary multi-objective optimization is that it can obtain non-dominated Pareto frontier solutions with respect to performance and shared weights. The experimental results show that we can reduce the neural network memory by $13.72 \sim14.98 \times$ on CIFAR-10, $11.61 \sim 12.99\times$ on CIFAR-100, and $7.44 \sim 8.58\times$ on ImageNet showcasing the effectiveness of the proposed deep neural network compression framework., Comment: 16 pages, 9 figures, submitted to IEEE Transactions on Neural Networks and Learning Systems

The study of spectral graph determination is a fascinating area of research in spectral graph theory and algebraic combinatorics. This field focuses on examining the spectral characterization of various classes of graphs, developing methods to construct or distinguish cospectral nonisomorphic graphs, and analyzing the conditions under which a graph's spectrum uniquely determines its structure. This paper presents an overview of both classical and recent advancements in these topics, along with newly obtained proofs of some existing results, which offer additional insights.

We derive analytic constraints on the weakly-coupled spectrum of theories with a massless scalar under the standard assumptions of the S-matrix bootstrap program. These bootstrap bounds apply to any theory (with or without gravity) with fully crossing symmetric (i.e. $stu$-symmetric) four-point amplitudes and generalize results for color- or flavor-ordered (i.e. $su$-symmetric) planar amplitudes recently proved by one of the authors. We assume that the theory is weakly-coupled below some cut-off, that the four-point massless scalar amplitude is polynomially-bounded in the Regge limit, and that this amplitude exchanges states with a discrete set of masses and a finite set of spins at each mass level. The spins and masses must then satisfy ``Sequential Spin Constraints" (SSC) and ``Sequential Mass Constraints" (SMC). The SSC requires the lightest spin-$j$ state to be lighter than the lightest spin-$(j+1)$ state (in the $su$-symmetric case) or the lightest spin-$(j+2)$ state (in the $stu$-symmetric case). The SMC requires the mass of the lightest spin-$j$ state to be smaller than some non-linear function of the masses of lower-spin states. Our results also apply to super-gluon and super-graviton amplitudes stripped of their polarization dependence. In particular, the open and closed superstring spectra saturate the SSC with maximum spins ${J_{n,\text{open}} = n+1}$ and ${J_{n,\text{closed}} = 2n+2}$, respectively, at the $n^\text{th}$ mass level., Comment: v1, 44 pages

At two-thirds the weight of aluminum, magnesium alloys have the potential to significantly reduce the fuel consumption of transportation vehicles. These advancements depend on our ability to optimize the desirable versus undesirable effects of deformation twins: three dimensional (3D) microstructural domains that form under mechanical stresses. Previously only characterized using surface or thin-film measurements, here, we present the first 3D in-situ characterization of deformation twinning inside an embedded grain over mesoscopic fields of view using dark-field X-ray microscopy supported by crystal plasticity finite element analysis. The results reveal the important role of triple junctions on twin nucleation, that twin growth behavior is irregular and can occur in several directions simultaneously, and that twin-grain and twin-twin junctions are the sites of localized dislocation accumulation, a necessary precursor to crack initiation.

Defects introduced to the surface of Bi(111) break the translational symmetry and modify the surface states locally. We present a theoretical and experimental study of the 2D defects on the surface of Bi(111) and the states that they induce. Bi crystals cleaved in ultrahigh vacuum (UHV) at low temperature (110 K) and the resulting ion-etched surface are investigated by low-energy electron diffraction (LEED), X-ray photoelectron spectroscopy, ultraviolet photoelectron spectroscopy (UPS), and scanning tunneling microscopy (STM) as well as spectroscopy (STS) techniques in combination with density functional theory (DFT) calculations. STS measurements of cleaved Bi(111) reveal that a commonly observed bilayer step edge has a lower density of states (DOS) around the Fermi level as compared to the atomic-flat terrace. Following ion bombardment, the Bi(111) surface reveals anomalous behavior at both 110 and 300 K: Surface periodicity is observed by LEED, and a significant increase in the number of bilayer step edges and energetically unfavorable monolayer steps is observed by STM. It is suggested that the newly exposed monolayer steps and the type A bilayer step edges result in an increase to the surface Fermi density as evidenced by UPS measurements and the Kohn-Sham DOS. These states appear to be thermodynamically stable under UHV conditions., Comment: 10 pages, 7 figures. See publisher's website for Supporting Information

We re-examine positivity bounds on the $2\to2$ scattering of identical massless real scalars with a novel perspective on how these bounds can be used to constrain the spectrum of UV theories. We propose that the entire space of consistent weakly-coupled (and generically non-supersymmetric) UV amplitudes is determined as a convex hull of the massive scalar amplitude and a one-parameter family of scalarless "extremal amplitudes" parameterized by the ratio of the masses of the two lightest massive states. Further, we propose that the extremal amplitudes can be constructed from a similar one-parameter set of maximally supersymmetric amplitudes, leading to the surprising possibility that the S-matrix bootstrap with maximal supersymmetry may be sufficient to determine the entire allowed space of four-point amplitudes! Finally, we show that minimal spectrum input reduces the allowed space of Wilson coefficients to small islands around the open string Dirac-Born-Infeld tree amplitude and the closed string Virasoro-Shapiro amplitude., Comment: 63 pages, 42 figures

Quantum imaging encompasses a broad range of methods that exploit the quantum properties of light to capture information about an object. One such approach involves using a two-photon quantum state, where only one photon interacts with the object being imaged while its entangled partner carries spatial or temporal information. To implement this technique, it is necessary to generate specific quantum states of light and detect photons at the single-photon level. While this method has been successfully demonstrated in the visible electromagnetic spectrum, extending it to X-rays has faced significant challenges due to the difficulties in producing a sufficient rate of X-ray photon pairs and detecting them with adequate resolution. Here, we demonstrate record high rates of correlated X-ray photon pairs produced via a spontaneous parametric down-conversion process and we employ these photons to perform quantum correlation imaging of several objects, including a biological sample (E. cardamomum seedpod). Notably, we report an unprecedented detection rate of about 6,300 pairs per hour and the observation of energy anti-correlation for the X-ray photon pairs. We also present a detailed analysis of the properties of the down-converted X-ray photons, as well as a comprehensive study of the correlation imaging formation, including a study of distortions and corrections. These results mark a substantial advancement in X-ray quantum imaging, expanding the possibilities of X-ray quantum optical technologies, and illustrating the pathway towards enhancing biological imaging with reduced radiation doses., Comment: 12 pages, 6 figures, publication preprint

This paper introduces a tuning-free method for both object insertion and subject-driven generation. The task involves composing an object, given multiple views, into a scene specified by either an image or text. Existing methods struggle to fully meet the task's challenging objectives: (i) seamlessly composing the object into the scene with photorealistic pose and lighting, and (ii) preserving the object's identity. We hypothesize that achieving these goals requires large scale supervision, but manually collecting sufficient data is simply too expensive. The key observation in this paper is that many mass-produced objects recur across multiple images of large unlabeled datasets, in different scenes, poses, and lighting conditions. We use this observation to create massive supervision by retrieving sets of diverse views of the same object. This powerful paired dataset enables us to train a straightforward text-to-image diffusion architecture to map the object and scene descriptions to the composited image. We compare our method, ObjectMate, with state-of-the-art methods for object insertion and subject-driven generation, using a single or multiple references. Empirically, ObjectMate achieves superior identity preservation and more photorealistic composition. Differently from many other multi-reference methods, ObjectMate does not require slow test-time tuning.

Targeting and personalization policies can be used to improve outcomes beyond the uniform policy that assigns the best performing treatment in an A/B test to everyone. Personalization relies on the presence of heterogeneity of treatment effects, yet, as we show in this paper, heterogeneity alone is not sufficient for personalization to be successful. We develop a statistical model to quantify "actionable heterogeneity," or the conditions when personalization is likely to outperform the best uniform policy. We show that actionable heterogeneity can be visualized as crossover interactions in outcomes across treatments and depends on three population-level parameters: within-treatment heterogeneity, cross-treatment correlation, and the variation in average responses. Our model can be used to predict the expected gain from personalization prior to running an experiment and also allows for sensitivity analysis, providing guidance on how changing treatments can affect the personalization gain. To validate our model, we apply five common personalization approaches to two large-scale field experiments with many interventions that encouraged flu vaccination. We find an 18% gain from personalization in one and a more modest 4% gain in the other, which is consistent with our model. Counterfactual analysis shows that this difference in the gains from personalization is driven by a drastic difference in within-treatment heterogeneity. However, reducing cross-treatment correlation holds a larger potential to further increase personalization gains. Our findings provide a framework for assessing the potential from personalization and offer practical recommendations for improving gains from targeting in multi-intervention settings.

Synthesizing a novel view from a single input image is a challenging task. Traditionally, this task was approached by estimating scene depth, warping, and inpainting, with machine learning models enabling parts of the pipeline. More recently, generative models are being increasingly employed in novel view synthesis (NVS), often encompassing the entire end-to-end system. In this work, we adapt a modern diffusion model architecture for end-to-end NVS in the pixel space, substantially outperforming previous state-of-the-art (SOTA) techniques. We explore different ways to encode geometric information into the network. Our experiments show that while these methods may enhance performance, their impact is minor compared to utilizing improved generative models. Moreover, we introduce a novel NVS training scheme that utilizes single-view datasets, capitalizing on their relative abundance compared to their multi-view counterparts. This leads to improved generalization capabilities to scenes with out-of-domain content.