Search

For non-empty sets X we define notions of distance and pseudo metric with values in a partially ordered set that has a smallest element $\theta $. If $h_X$ is a distance in $X$ (respectively, a pseudo metric in $X$), then the pair $(X,h_X)$ is called a distance (respectively, a pseudo metric) space. If $(T,h_T)$ and $(X,h_X)$ are pseudo metric spaces, $(Y,h_Y)$ is a distance space, and $H(T,X)$ is a class of Lipschitz mappings $f\colon T\to X$, for a broad family of mappings $\Lambda\colon H (T,X)\to Y$, we obtain a sharp inequality that estimates the deviation $h_Y(\Lambda f(\cdot),\Lambda f(t))$ in terms of the function $h_T(\cdot, t)$. We also show that many known estimates of such kind are contained in our general result.

Microscopic interpretation of histopathology images underlies many important diagnostic and treatment decisions. While advances in vision-language modeling raise new opportunities for analysis of such images, the gigapixel-scale size of whole slide images (WSIs) introduces unique challenges. Additionally, pathology reports simultaneously highlight key findings from small regions while also aggregating interpretation across multiple slides, often making it difficult to create robust image-text pairs. As such, pathology reports remain a largely untapped source of supervision in computational pathology, with most efforts relying on region-of-interest annotations or self-supervision at the patch-level. In this work, we develop a vision-language model based on the BLIP-2 framework using WSIs paired with curated text from pathology reports. This enables applications utilizing a shared image-text embedding space, such as text or image retrieval for finding cases of interest, as well as integration of the WSI encoder with a frozen large language model (LLM) for WSI-based generative text capabilities such as report generation or AI-in-the-loop interactions. We utilize a de-identified dataset of over 350,000 WSIs and diagnostic text pairs, spanning a wide range of diagnoses, procedure types, and tissue types. We present pathologist evaluation of text generation and text retrieval using WSI embeddings, as well as results for WSI classification and workflow prioritization (slide-level triaging). Model-generated text for WSIs was rated by pathologists as accurate, without clinically significant error or omission, for 78% of WSIs on average. This work demonstrates exciting potential capabilities for language-aligned WSI embeddings., Comment: 9 main pages and 19 pages of supplemental material; 3 main tables, 3 main figures and 11 supplemental tables, 7 supplemental figures

Benchmarks that closely reflect downstream application scenarios are essential for the streamlined adoption of new research in tabular machine learning (ML). In this work, we examine existing tabular benchmarks and find two common characteristics of industry-grade tabular data that are underrepresented in the datasets available to the academic community. First, tabular data often changes over time in real-world deployment scenarios. This impacts model performance and requires time-based train and test splits for correct model evaluation. Yet, existing academic tabular datasets often lack timestamp metadata to enable such evaluation. Second, a considerable portion of datasets in production settings stem from extensive data acquisition and feature engineering pipelines. For each specific dataset, this can have a different impact on the absolute and relative number of predictive, uninformative, and correlated features, which in turn can affect model selection. To fill the aforementioned gaps in academic benchmarks, we introduce TabReD -- a collection of eight industry-grade tabular datasets covering a wide range of domains from finance to food delivery services. We assess a large number of tabular ML models in the feature-rich, temporally-evolving data setting facilitated by TabReD. We demonstrate that evaluation on time-based data splits leads to different methods ranking, compared to evaluation on random splits more common in academic benchmarks. Furthermore, on the TabReD datasets, MLP-like architectures and GBDT show the best results, while more sophisticated DL models are yet to prove their effectiveness., Comment: Code: https://github.com/yandex-research/tabred (V2: fix the link to the code in this comment; no changes to the PDF)

Diffusion distillation represents a highly promising direction for achieving faithful text-to-image generation in a few sampling steps. However, despite recent successes, existing distilled models still do not provide the full spectrum of diffusion abilities, such as real image inversion, which enables many precise image manipulation methods. This work aims to enrich distilled text-to-image diffusion models with the ability to effectively encode real images into their latent space. To this end, we introduce invertible Consistency Distillation (iCD), a generalized consistency distillation framework that facilitates both high-quality image synthesis and accurate image encoding in only 3-4 inference steps. Though the inversion problem for text-to-image diffusion models gets exacerbated by high classifier-free guidance scales, we notice that dynamic guidance significantly reduces reconstruction errors without noticeable degradation in generation performance. As a result, we demonstrate that iCD equipped with dynamic guidance may serve as a highly effective tool for zero-shot text-guided image editing, competing with more expensive state-of-the-art alternatives., Comment: Project page: https://yandex-research.github.io/invertible-cd/

The field of artificial intelligence (AI) is rapidly influencing health and healthcare, but bias and poor performance persists for populations who face widespread structural oppression. Previous work has clearly outlined the need for more rigorous attention to data representativeness and model performance to advance equity and reduce bias. However, there is an opportunity to also improve the explainability of AI by leveraging best practices of social epidemiology and health equity to help us develop hypotheses for associations found. In this paper, we focus on explainable AI (XAI) and describe a framework for interdisciplinary expert panel review to discuss and critically assess AI model explanations from multiple perspectives and identify areas of bias and directions for future research. We emphasize the importance of the interdisciplinary expert panel to produce more accurate, equitable interpretations which are historically and contextually informed. Interdisciplinary panel discussions can help reduce bias, identify potential confounders, and identify opportunities for additional research where there are gaps in the literature. In turn, these insights can suggest opportunities for AI model improvement.

In the rapidly progressing field of generative models, the development of efficient and high-fidelity text-to-image diffusion systems represents a significant frontier. This study introduces YaART, a novel production-grade text-to-image cascaded diffusion model aligned to human preferences using Reinforcement Learning from Human Feedback (RLHF). During the development of YaART, we especially focus on the choices of the model and training dataset sizes, the aspects that were not systematically investigated for text-to-image cascaded diffusion models before. In particular, we comprehensively analyze how these choices affect both the efficiency of the training process and the quality of the generated images, which are highly important in practice. Furthermore, we demonstrate that models trained on smaller datasets of higher-quality images can successfully compete with those trained on larger datasets, establishing a more efficient scenario of diffusion models training. From the quality perspective, YaART is consistently preferred by users over many existing state-of-the-art models., Comment: Prompts and additional information are available on the project page, see https://ya.ru/ai/art/paper-yaart-v1

This paper introduces a new data-driven, non-parametric method for image quality and aesthetics assessment, surpassing existing approaches and requiring no prompt engineering or fine-tuning. We eliminate the need for expressive textual embeddings by proposing efficient image anchors in the data. Through extensive evaluations of 7 state-of-the-art self-supervised models, our method demonstrates superior performance and robustness across various datasets and benchmarks. Notably, it achieves high agreement with human assessments even with limited data and shows high robustness to the nature of data and their pre-processing pipeline. Our contributions offer a streamlined solution for assessment of images while providing insights into the perception of visual information.

The emergence of accurate open large language models (LLMs) has led to a race towards performant quantization techniques which can enable their execution on end-user devices. In this paper, we revisit the problem of "extreme" LLM compression-defined as targeting extremely low bit counts, such as 2 to 3 bits per parameter-from the point of view of classic methods in Multi-Codebook Quantization (MCQ). Our algorithm, called AQLM, generalizes the classic Additive Quantization (AQ) approach for information retrieval to advance the state-of-the-art in LLM compression, via two innovations: 1) learned additive quantization of weight matrices in input-adaptive fashion, and 2) joint optimization of codebook parameters across each transformer blocks. Broadly, AQLM is the first scheme that is Pareto optimal in terms of accuracy-vs-model-size when compressing to less than 3 bits per parameter, and significantly improves upon all known schemes in the extreme compression (2bit) regime. In addition, AQLM is practical: we provide fast GPU and CPU implementations of AQLM for token generation, which enable us to match or outperform optimized FP16 implementations for speed, while executing in a much smaller memory footprint., Comment: ICML, 2024

Knowledge distillation methods have recently shown to be a promising direction to speedup the synthesis of large-scale diffusion models by requiring only a few inference steps. While several powerful distillation methods were recently proposed, the overall quality of student samples is typically lower compared to the teacher ones, which hinders their practical usage. In this work, we investigate the relative quality of samples produced by the teacher text-to-image diffusion model and its distilled student version. As our main empirical finding, we discover that a noticeable portion of student samples exhibit superior fidelity compared to the teacher ones, despite the "approximate" nature of the student. Based on this finding, we propose an adaptive collaboration between student and teacher diffusion models for effective text-to-image synthesis. Specifically, the distilled model produces the initial sample, and then an oracle decides whether it needs further improvements with a slow teacher model. Extensive experiments demonstrate that the designed pipeline surpasses state-of-the-art text-to-image alternatives for various inference budgets in terms of human preference. Furthermore, the proposed approach can be naturally used in popular applications such as text-guided image editing and controllable generation., Comment: CVPR2024 camera ready v2

Deep learning (DL) models for tabular data problems (e.g. classification, regression) are currently receiving increasingly more attention from researchers. However, despite the recent efforts, the non-DL algorithms based on gradient-boosted decision trees (GBDT) remain a strong go-to solution for these problems. One of the research directions aimed at improving the position of tabular DL involves designing so-called retrieval-augmented models. For a target object, such models retrieve other objects (e.g. the nearest neighbors) from the available training data and use their features and labels to make a better prediction. In this work, we present TabR -- essentially, a feed-forward network with a custom k-Nearest-Neighbors-like component in the middle. On a set of public benchmarks with datasets up to several million objects, TabR marks a big step forward for tabular DL: it demonstrates the best average performance among tabular DL models, becomes the new state-of-the-art on several datasets, and even outperforms GBDT models on the recently proposed "GBDT-friendly" benchmark (see Figure 1). Among the important findings and technical details powering TabR, the main ones lie in the attention-like mechanism that is responsible for retrieving the nearest neighbors and extracting valuable signal from them. In addition to the much higher performance, TabR is simple and significantly more efficient compared to prior retrieval-based tabular DL models., Comment: Code: https://github.com/yandex-research/tabular-dl-tabr

For a function $f$ from the Sobolev space $W^{1,p}(C)$ ($C\subset\mathbb{R}^d$ is an open convex cone), a sharp inequality that estimates $\| f\|_{L_{\infty}}$ via the $L_{p}$-norm of its gradient and a seminorm of the function is obtained. With the help of this inequality, a sharp inequality is proved, which estimates the ${L_{\infty}}$-norm of the Radon--Nikodym derivative of a charge defined on Lebesgue measurable subsets of $C$ via the $L_p$-norm of the gradient of this derivative and a seminorm of the charge. In the case, when $C=\mathbb{R}_+^m\times \mathbb{R}^{d-m}$, $0\le m\le d$, we obtain inequalities that estimate the ${L_{\infty}}$-norm of a mixed derivative of a function $f\colon C\to \mathbb{R}$ using its ${L_{\infty}}$-norm and the $L_p$-norm of the gradient of the function's mixed derivative.

We obtain a sharp Nagy type inequality in a metric space $(X,\rho)$ with measure $\mu$ that estimates the uniform norm of a function using its $\|\cdot\|_{H^\omega}$ -- norm determined by a modulus of continuity $\omega$, and a seminorm that is defined on a space of locally integrable functions. We consider charges $\nu$ that are defined on the set of $\mu$-measurable subsets of $X$ and are absolutely continuous with respect to $\mu$. Using the obtained Nagy type inequality, we prove a sharp Landau-Kolmogorov type inequality that estimates the uniform norm of a Radon-Nikodym derivative of a charge via a $\|\cdot\|_{H^\omega}$-norm of this derivative, and a seminorm defined on the space of such charges. We also prove a sharp inequality for a hypersingular integral operator. In the case $X=\mathbb{R}_+^m\times \mathbb{R}^{d-m}$, $0\le m\le d$, we obtain inequalities that estimate the uniform norm of a mixed derivative of a function using the uniform norm of the function and the $\|\cdot\|_{H^\omega}$-norm of its mixed derivative.

AI models have shown promise in many medical imaging tasks. However, our ability to explain what signals these models have learned is severely lacking. Explanations are needed in order to increase the trust in AI-based models, and could enable novel scientific discovery by uncovering signals in the data that are not yet known to experts. In this paper, we present a method for automatic visual explanations leveraging team-based expertise by generating hypotheses of what visual signals in the images are correlated with the task. We propose the following 4 steps: (i) Train a classifier to perform a given task (ii) Train a classifier guided StyleGAN-based image generator (StylEx) (iii) Automatically detect and visualize the top visual attributes that the classifier is sensitive towards (iv) Formulate hypotheses for the underlying mechanisms, to stimulate future research. Specifically, we present the discovered attributes to an interdisciplinary panel of experts so that hypotheses can account for social and structural determinants of health. We demonstrate results on eight prediction tasks across three medical imaging modalities: retinal fundus photographs, external eye photographs, and chest radiographs. We showcase examples of attributes that capture clinically known features, confounders that arise from factors beyond physiological mechanisms, and reveal a number of physiologically plausible novel attributes. Our approach has the potential to enable researchers to better understand, improve their assessment, and extract new knowledge from AI-based models. Importantly, we highlight that attributes generated by our framework can capture phenomena beyond physiology or pathophysiology, reflecting the real world nature of healthcare delivery and socio-cultural factors. Finally, we intend to release code to enable researchers to train their own StylEx models and analyze their predictive tasks., Comment: 43 pages, 1 figure

In this article we prove sharp Landau--Kolmogorov type inequalities on a class of charges defined on Lebesgue measurable subsets of a cone in $\mathbb{R}^d$, $d\geq 1$, that are absolutely continuous with respect to the Lebesgue measure. In addition we solve the Stechkin problem of approximation of the Radon--Nikodym derivative of such charges by bounded operators and two related problems. As an application, we also solve these extremal problems on classes of essentially bounded functions $f$ such that their distributional partial derivative $\frac{\partial ^d f}{\partial x_1\ldots\partial x_d}$ belongs to the Sobolev space $W^{1,\infty}$.

Recent advances in diffusion models enable many powerful instruments for image editing. One of these instruments is text-driven image manipulations: editing semantic attributes of an image according to the provided text description. % Popular text-conditional diffusion models offer various high-quality image manipulation methods for a broad range of text prompts. Existing diffusion-based methods already achieve high-quality image manipulations for a broad range of text prompts. However, in practice, these methods require high computation costs even with a high-end GPU. This greatly limits potential real-world applications of diffusion-based image editing, especially when running on user devices. In this paper, we address efficiency of the recent text-driven editing methods based on unconditional diffusion models and develop a novel algorithm that learns image manipulations 4.5-10 times faster and applies them 8 times faster. We carefully evaluate the visual quality and expressiveness of our approach on multiple datasets using human annotators. Our experiments demonstrate that our algorithm achieves the quality of much more expensive methods. Finally, we show that our approach can adapt the pretrained model to the user-specified image and text description on the fly just for 4 seconds. In this setting, we notice that more compact unconditional diffusion models can be considered as a rational alternative to the popular text-conditional counterparts.

Let $G$ be an undirected network with a distinguished set of terminals $T \subseteq V(G)$ and edge capacities $cap: E(G) \rightarrow \mathbb{R}_+$. By an odd $T$-walk we mean a walk in $G$ (with possible vertex and edge self-intersections) connecting two distinct terminals and consisting of an odd number of edges. Inspired by the work of Schrijver and Seymour on odd path packing for two terminals, we consider packings of odd $T$-walks subject to capacities $cap$. First, we present a strongly polynomial time algorithm for constructing a maximum fractional packing of odd $T$-walks. For even integer capacities, our algorithm constructs a packing that is half-integer. Additionally, if $cap(\delta(v))$ is divisible by 4 for any $v \in V(G) - T$, our algorithm constructs an integer packing. Second, we establish and prove the corresponding min-max relation. Third, if $G$ is inner Eulerian (i.e. degrees of all nodes in $V(G) - T$ are even) and $cap(e) = 2$ for all $e \in E$, we show that there exists an integer packing of odd $T$-trails (i.e. odd $T$-walks with no repeated edges) of the same value as in case of odd $T$-walks, and this packing can be found in polynomial time. To achieve the above goals, we establish a connection between packings of odd $T$-walks and $T$-trails and certain multiflow problems in undirected and bidirected graphs., Comment: 20 pages, 7 figures, accepted to STACS 2023

In reliable decision-making systems based on machine learning, models have to be robust to distributional shifts or provide the uncertainty of their predictions. In node-level problems of graph learning, distributional shifts can be especially complex since the samples are interdependent. To evaluate the performance of graph models, it is important to test them on diverse and meaningful distributional shifts. However, most graph benchmarks considering distributional shifts for node-level problems focus mainly on node features, while structural properties are also essential for graph problems. In this work, we propose a general approach for inducing diverse distributional shifts based on graph structure. We use this approach to create data splits according to several structural node properties: popularity, locality, and density. In our experiments, we thoroughly evaluate the proposed distributional shifts and show that they can be quite challenging for existing graph models. We also reveal that simple models often outperform more sophisticated methods on the considered structural shifts. Finally, our experiments provide evidence that there is a trade-off between the quality of learned representations for the base classification task under structural distributional shift and the ability to separate the nodes from different distributions using these representations.

Node classification is a classical graph machine learning task on which Graph Neural Networks (GNNs) have recently achieved strong results. However, it is often believed that standard GNNs only work well for homophilous graphs, i.e., graphs where edges tend to connect nodes of the same class. Graphs without this property are called heterophilous, and it is typically assumed that specialized methods are required to achieve strong performance on such graphs. In this work, we challenge this assumption. First, we show that the standard datasets used for evaluating heterophily-specific models have serious drawbacks, making results obtained by using them unreliable. The most significant of these drawbacks is the presence of a large number of duplicate nodes in the datasets Squirrel and Chameleon, which leads to train-test data leakage. We show that removing duplicate nodes strongly affects GNN performance on these datasets. Then, we propose a set of heterophilous graphs of varying properties that we believe can serve as a better benchmark for evaluating the performance of GNNs under heterophily. We show that standard GNNs achieve strong results on these heterophilous graphs, almost always outperforming specialized models. Our datasets and the code for reproducing our experiments are available at https://github.com/yandex-research/heterophilous-graphs

Text-to-image generation models represent the next step of evolution in image synthesis, offering a natural way to achieve flexible yet fine-grained control over the result. One emerging area of research is the fast adaptation of large text-to-image models to smaller datasets or new visual concepts. However, many efficient methods of adaptation have a long training time, which limits their practical applications, slows down experiments, and spends excessive GPU resources. In this work, we study the training dynamics of popular text-to-image personalization methods (such as Textual Inversion or DreamBooth), aiming to speed them up. We observe that most concepts are learned at early stages and do not improve in quality later, but standard training convergence metrics fail to indicate that. Instead, we propose a simple drop-in early stopping criterion that only requires computing the regular training objective on a fixed set of inputs for all training iterations. Our experiments on Stable Diffusion for 48 different concepts and three personalization methods demonstrate the competitive performance of our approach, which makes adaptation up to 8 times faster with no significant drops in quality., Comment: Accepted to Conference on Neural Information Processing Systems (NeurIPS) 2023. 20 pages, 15 figures. Code: https://github.com/yandex-research/DVAR