Search

Scientists and engineers use simulators to model empirically observed phenomena. However, tuning the parameters of a simulator to ensure its outputs match observed data presents a significant challenge. Simulation-based inference (SBI) addresses this by enabling Bayesian inference for simulators, identifying parameters that match observed data and align with prior knowledge. Unlike traditional Bayesian inference, SBI only needs access to simulations from the model and does not require evaluations of the likelihood-function. In addition, SBI algorithms do not require gradients through the simulator, allow for massive parallelization of simulations, and can perform inference for different observations without further simulations or training, thereby amortizing inference. Over the past years, we have developed, maintained, and extended $\texttt{sbi}$, a PyTorch-based package that implements Bayesian SBI algorithms based on neural networks. The $\texttt{sbi}$ toolkit implements a wide range of inference methods, neural network architectures, sampling methods, and diagnostic tools. In addition, it provides well-tested default settings but also offers flexibility to fully customize every step of the simulation-based inference workflow. Taken together, the $\texttt{sbi}$ toolkit enables scientists and engineers to apply state-of-the-art SBI methods to black-box simulators, opening up new possibilities for aligning simulations with empirically observed data.

Present-day software development faces three major challenges: complexity, time consumption, and high costs. Developing large software systems often requires battalions of teams and considerable time for meetings, which end without any action, resulting in unproductive cycles, delayed progress, and increased cost. What if, instead of large meetings with no immediate results, the software product is completed by the end of the meeting? In response, we present a vision for a system called TimeLess, designed to reshape the software development process by enabling immediate action during meetings. The goal is to shift meetings from planning discussions to productive, action-oriented sessions. This approach minimizes the time and effort required for development, allowing teams to focus on critical decision-making while AI agents execute development tasks based on the meeting discussions. We will employ multiple AI agents that work collaboratively to capture human discussions and execute development tasks in real time. This represents a step toward next-generation software development environments, where human expertise drives strategy and AI accelerates task execution., Comment: 5 pages, 4 figure, and 1 table

We report on the electronic, structural, and optical properties of epitaxial GaSbBi films with varying Bi-concentration (up to 7%Bi) grown on semi-insulating GaAs(100) substrates. The 1 $\mu$m thick GaSbBi epilayers exhibit fully relaxed narrow X-ray diffraction peaks and smooth surface morphology comparable to that of high-quality GaSb epilayers on GaAs. Low temperature photoluminescence spectra exhibit band gap shrinkage consistent with Bi alloying. Electrical Hall measurements indicate reduction of hole concentration and no change in the hole mobilities with increasing Bi content for the nominally undoped GaSbBi alloy. The residual hole concentration reduces from $10^{18} cm^{-3}$ level for a reference GaSb sample, to low $10^{17} cm^{-3}$ level with increasing Bi content. Hole mobility values of around $300 cm^{2}/Vs$ are observed independent of the Bi content. These dependencies are attributed to the Bi surfactant effect and Bi-induced defect formation., Comment: 11 pages, 4 figures

The discovery and identification of molecules in biological and environmental samples is crucial for advancing biomedical and chemical sciences. Tandem mass spectrometry (MS/MS) is the leading technique for high-throughput elucidation of molecular structures. However, decoding a molecular structure from its mass spectrum is exceptionally challenging, even when performed by human experts. As a result, the vast majority of acquired MS/MS spectra remain uninterpreted, thereby limiting our understanding of the underlying (bio)chemical processes. Despite decades of progress in machine learning applications for predicting molecular structures from MS/MS spectra, the development of new methods is severely hindered by the lack of standard datasets and evaluation protocols. To address this problem, we propose MassSpecGym -- the first comprehensive benchmark for the discovery and identification of molecules from MS/MS data. Our benchmark comprises the largest publicly available collection of high-quality labeled MS/MS spectra and defines three MS/MS annotation challenges: \textit{de novo} molecular structure generation, molecule retrieval, and spectrum simulation. It includes new evaluation metrics and a generalization-demanding data split, therefore standardizing the MS/MS annotation tasks and rendering the problem accessible to the broad machine learning community. MassSpecGym is publicly available at \url{https://github.com/pluskal-lab/MassSpecGym}.

Consider a Jordan curve on the Riemann sphere passing through $n \ge 3$ given points. We show that in each relative isotopy class of such curves, there exists a unique curve that minimizes the Loewner energy. These curves have the property that each arc between two consecutive points is a hyperbolic geodesic in the domain bounded by the other arcs. This geodesic property lets us define a complex projective structure on the punctured sphere. We show that its associated quadratic differential has simple poles, whose residues (accessory parameters) are given by the Wirtinger derivatives of the minimal Loewner energy, a result resembling Polyakov's conjecture for the Fuchsian projective structure that was later proven by Takhtajan and Zograf. Finally, we show that the projective structures we obtain are related to the Fuchsian projective structures via a $\pi$-grafting., Comment: 37 pages, 7 figures

Current speech deepfake detection approaches perform satisfactorily against known adversaries; however, generalization to unseen attacks remains an open challenge. The proliferation of speech deepfakes on social media underscores the need for systems that can generalize to unseen attacks not observed during training. We address this problem from the perspective of meta-learning, aiming to learn attack-invariant features to adapt to unseen attacks with very few samples available. This approach is promising since generating of a high-scale training dataset is often expensive or infeasible. Our experiments demonstrated an improvement in the Equal Error Rate (EER) from 21.67% to 10.42% on the InTheWild dataset, using just 96 samples from the unseen dataset. Continuous few-shot adaptation ensures that the system remains up-to-date., Comment: 6 pages, accepted to the IEEE Spoken Language Technology Workshop (SLT) 2024

Using unique research data, we investigate disability retirement risk under the statutory public sector pension scheme in Finland. The statistical analysis yields two indicators: risk for upcoming permanent disability pension and critical duration of sickness absence days for public sector occupations. Statistical analysis is based on logistic regression model where the outcome is the disability pension, using sickness benefit spells and other individual background information as covariates. The results underline the importance of minimizing the sickness spells and their duration to the risk and reveal differences in risk across occupations. We conclude that the proposed risk model is a promising tool which can help employers and the pension industry in preventing permanent disability., Comment: 16 pages, 2 figures and 6 tables

Despite the recent progress, existing frame interpolation methods still struggle with processing extremely high resolution input and handling challenging cases such as repetitive textures, thin objects, and large motion. To address these issues, we introduce a patch-based cascaded pixel diffusion model for frame interpolation, HiFI, that excels in these scenarios while achieving competitive performance on standard benchmarks. Cascades, which generate a series of images from low- to high-resolution, can help significantly with large or complex motion that require both global context for a coarse solution and detailed context for high resolution output. However, contrary to prior work on cascaded diffusion models which perform diffusion on increasingly large resolutions, we use a single model that always performs diffusion at the same resolution and upsamples by processing patches of the inputs and the prior solution. We show that this technique drastically reduces memory usage at inference time and also allows us to use a single model at test time, solving both frame interpolation and spatial up-sampling, saving training cost. We show that HiFI helps significantly with high resolution and complex repeated textures that require global context. HiFI demonstrates comparable or beyond state-of-the-art performance on multiple benchmarks (Vimeo, Xiph, X-Test, SEPE-8K). On our newly introduced dataset that focuses on particularly challenging cases, HiFI also significantly outperforms other baselines on these cases. Please visit our project page for video results: https://hifi-diffusion.github.io, Comment: Project page: https://hifi-diffusion.github.io/

Nowadays, Mixed Reality (MR) headsets are a game-changer for knowledge work. Unlike stationary monitors, MR headsets allow users to work with large virtual displays anywhere they wear the headset, whether in a professional office, a public setting like a cafe, or a quiet space like a library. This study compares four different layouts (eye level-close, eye level-far, below eye level-close, below eye level-far) of virtual displays regarding feelings of safety, perceived productivity, and social acceptability when working with MR in public. We test which layout is most preferred by users and seek to understand which factors affect users' layout preferences. The aim is to derive useful insights for designing better MR layouts. A field study in a public library was conducted using a within-subject design. While the participants interact with a layout, they are asked to work on a planning task. The results from a repeated measure ANOVA show a statistically significant effect on productivity but not on safety and social acceptability. Additionally, we report preferences expressed by the users regarding the layouts and using MR in public., Comment: Part of the 23rd annual IEEE International Symposium on Mixed and Augmented Reality (ISMAR 2024)

Large power consumption of silicon CMOS electronics is a challenge in very-large-scale integrated circuits and a major roadblock to fault-tolerant quantum computation. Matching the power dissipation of Si-MOSFETs to the thermal budget at deep cryogenic temperatures, below 1 K, requires switching performance beyond levels facilitated by currently available CMOS technologies. We have manufactured fully depleted silicon-on-insulator MOSFETs tailored for overcoming the power dissipation barrier towards sub-1 K applications. With these cryo-optimized transistors we achieve a major milestone of reaching subthreshold swing of 0.3 mV/dec at 420 mK, thereby enabling very-large-scale integration of cryo-CMOS electronics for ultra-low temperature applications.

Quantum computing has tremendous potential to overcome some of the fundamental limitations present in classical information processing. Yet, today's technological limitations in the quality and scaling prevent exploiting its full potential. Quantum computing based on superconducting quantum processing units (QPUs) is among the most promising approaches towards practical quantum advantage. In this article the basic technological approach of IQM Quantum Computers is described covering both the QPU and the rest of the full-stack quantum computer. In particular, the focus is on a 20-qubit quantum computer featuring the Garnet QPU and its architecture, which we will scale up to 150 qubits. We also present QPU and system-level benchmarks, including a median 2-qubit gate fidelity of 99.5% and genuinely entangling all 20 qubits in a Greenberger-Horne-Zeilinger (GHZ) state.

We introduce Seed-Music, a suite of music generation systems capable of producing high-quality music with fine-grained style control. Our unified framework leverages both auto-regressive language modeling and diffusion approaches to support two key music creation workflows: controlled music generation and post-production editing. For controlled music generation, our system enables vocal music generation with performance controls from multi-modal inputs, including style descriptions, audio references, musical scores, and voice prompts. For post-production editing, it offers interactive tools for editing lyrics and vocal melodies directly in the generated audio. We encourage readers to listen to demo audio examples at https://team.doubao.com/seed-music "https://team.doubao.com/seed-music"., Comment: Seed-Music technical report, 20 pages, 5 figures

Progress in the task of symbolic music generation may be lagging behind other tasks like audio and text generation, in part because of the scarcity of symbolic training data. In this paper, we leverage the greater scale of audio music data by applying pre-trained MIR models (for transcription, beat tracking, structure analysis, etc.) to extract symbolic events and encode them into token sequences. To the best of our knowledge, this work is the first to demonstrate the feasibility of training symbolic generation models solely from auto-transcribed audio data. Furthermore, to enhance the controllability of the trained model, we introduce SymPAC (Symbolic Music Language Model with Prompting And Constrained Generation), which is distinguished by using (a) prompt bars in encoding and (b) a technique called Constrained Generation via Finite State Machines (FSMs) during inference time. We show the flexibility and controllability of this approach, which may be critical in making music AI useful to creators and users., Comment: ISMIR 2024

We introduce an information theoretic criterion for Bayesian network structure learning which we call quotient normalized maximum likelihood (qNML). In contrast to the closely related factorized normalized maximum likelihood criterion, qNML satisfies the property of score equivalence. It is also decomposable and completely free of adjustable hyperparameters. For practical computations, we identify a remarkably accurate approximation proposed earlier by Szpankowski and Weinberger. Experiments on both simulated and real data demonstrate that the new criterion leads to parsimonious models with good predictive accuracy., Comment: Accepted to AISTATS 2018

Recent research has shown that integrating artificial intelligence (AI) into wireless communication systems can significantly improve spectral efficiency. However, the prevalent use of simulated radio channel data for training and validating neural network-based radios raises concerns about their generalization capability to diverse real-world environments. To address this, we conducted empirical over-the-air (OTA) experiments using software-defined radio (SDR) technology to test the performance of an NN-based orthogonal frequency division multiplexing (OFDM) receiver in a real-world small cell scenario. Our assessment reveals that the performance of receivers trained on diverse 3GPP TS38.901 channel models and broad parameter ranges significantly surpasses conventional receivers in our testing environment, demonstrating strong generalization to a new environment. Conversely, setting simulation parameters to narrowly reflect the actual measurement environment led to suboptimal OTA performance, highlighting the crucial role of rich and randomized training data in improving the NN-based receiver's performance. While our empirical test results are promising, they also suggest that developing new channel models tailored for training these learned receivers would enhance their generalization capability and reduce training time. Our testing was limited to a relatively narrow environment, and we encourage further testing in more complex environments.

Given a non-negative real sequence $\{c_n\}_n$ such that the series $\sum_{n=1}^{\infty}c_n$ diverges, it is known that the size of an infinite subset $A\subset\mathbb{N}$ can be measured in terms of the linear density such that the sub-series $\sum_{n\in A}c_n$ either (a) converges or (b) still diverges. The purpose of this research is to study these convergence/divergence questions by measuring the size of the set $A\subset\mathbb{N}$ in a more precise way in terms of the recently introduced asymptotic $\psi$-density. The convergence of the associated sub-signed series $\sum_{n=1 }^{\infty}m_nc_n$ is also discussed, where $\{m_n\}_n$ is a real sequence with values restricted to the set $\{-1, 0, 1\}$., Comment: 27 pages

Toxicity classification for voice heavily relies on the semantic content of speech. We propose a novel framework that utilizes cross-modal learning to integrate the semantic embedding of text into a multilabel speech toxicity classifier during training. This enables us to incorporate textual information during training while still requiring only audio during inference. We evaluate this classifier on large-scale datasets with real-world characteristics to validate the effectiveness of this framework. Through ablation studies, we demonstrate that general-purpose semantic text embeddings are rich and aligned with speech for toxicity classification purposes. Conducting experiments across multiple languages at scale, we show improvements in voice toxicity classification across five languages and different toxicity categories., Comment: Accepted to INTERSPEECH 2024

We propose ExtraNeRF, a novel method for extrapolating the range of views handled by a Neural Radiance Field (NeRF). Our main idea is to leverage NeRFs to model scene-specific, fine-grained details, while capitalizing on diffusion models to extrapolate beyond our observed data. A key ingredient is to track visibility to determine what portions of the scene have not been observed, and focus on reconstructing those regions consistently with diffusion models. Our primary contributions include a visibility-aware diffusion-based inpainting module that is fine-tuned on the input imagery, yielding an initial NeRF with moderate quality (often blurry) inpainted regions, followed by a second diffusion model trained on the input imagery to consistently enhance, notably sharpen, the inpainted imagery from the first pass. We demonstrate high-quality results, extrapolating beyond a small number of (typically six or fewer) input views, effectively outpainting the NeRF as well as inpainting newly disoccluded regions inside the original viewing volume. We compare with related work both quantitatively and qualitatively and show significant gains over prior art., Comment: 8 pages, 8 figures, CVPR2024

This article rethinks the role of arts in STEAM education, emphasizing its importance in AI literacy within K-12 contexts. Arguing against the marginalization of arts, the paper is structured around four key domains: language studies, philosophy, social studies, and visual arts. Each section addresses critical AI-related phenomena and provides pedagogical strate-gies for effective integration into STEAM education. Language studies focus on media representations and the probabilistic nature of AI language models. The philosophy section examines anthropomorphism, ethics, and the misconstrued human-like capabilities of AI. Social studies discuss AI's societal impacts, biases, and ethical considerations in data prac-tices. Visual arts explore the implications of generative AI on artistic processes and intellec-tual property. The article concludes by advocating for a robust inclusion of arts in STEAM to foster a holistic, equitable, and sustainable understanding of AI, ultimately inspiring technologies that promote fairness and creativity.

In prior work, Gupta et al. (SPAA 2022) presented a distributed algorithm for multiplying sparse $n \times n$ matrices, using $n$ computers. They assumed that the input matrices are uniformly sparse--there are at most $d$ non-zeros in each row and column--and the task is to compute a uniformly sparse part of the product matrix. The sparsity structure is globally known in advance (this is the supported setting). As input, each computer receives one row of each input matrix, and each computer needs to output one row of the product matrix. In each communication round each computer can send and receive one $O(\log n)$-bit message. Their algorithm solves this task in $O(d^{1.907})$ rounds, while the trivial bound is $O(d^2)$. We improve on the prior work in two dimensions: First, we show that we can solve the same task faster, in only $O(d^{1.832})$ rounds. Second, we explore what happens when matrices are not uniformly sparse. We consider the following alternative notions of sparsity: row-sparse matrices (at most $d$ non-zeros per row), column-sparse matrices, matrices with bounded degeneracy (we can recursively delete a row or column with at most $d$ non-zeros), average-sparse matrices (at most $dn$ non-zeros in total), and general matrices.

Atomic ensembles strongly interacting with light constitute rich quantum-optical many-body systems, with the potential for observing cooperative effects and dissipative nonequilibrium phase transitions. We theoretically analyze the conditions under which a driven atomic ensemble in free space, characterized by strong dipole-dipole interactions and large spatial extent, can undergo a superradiant phase transition, also known as cooperative resonance fluorescence. In an atomic array, stationary states that conserve the collective pseudospin exhibit completely cooperative decay and undergo a second-order phase transition in the large atom number limit. In contrast, decay mechanisms on longer timescales that fail to conserve pseudospin can lead to discontinuous first-order phase transition at a critical finite atom number, disrupting cooperation despite sharing many similar observable characteristics. A hallmark of the superradiant phase transition is an abrupt shift from total light reflection off the atoms to rapidly increasing transmission, accompanied by significant quantum fluctuations, as a function of light intensity.

Differential abundance analysis is a key component of microbiome studies. While dozens of methods for it exist, currently, there is no consensus on the preferred methods. Correctness of results in differential abundance analysis is an ambiguous concept that cannot be evaluated without employing simulated data, but we argue that consistency of results across datasets should be considered as an essential quality of a well-performing method. We compared the performance of 14 differential abundance analysis methods employing datasets from 54 taxonomic profiling studies based on 16S rRNA gene or shotgun sequencing. For each method, we examined how the results replicated between random partitions of each dataset and between datasets from independent studies. While certain methods showed good consistency, some widely used methods were observed to produce a substantial number of conflicting findings. Overall, the highest consistency without unnecessary reduction in sensitivity was attained by analyzing relative abundances with a non-parametric method (Wilcoxon test or ordinal regression model) or linear regression (MaAsLin2). Comparable performance was also attained by analyzing presence/absence of taxa with logistic regression.

We present VIDIM, a generative model for video interpolation, which creates short videos given a start and end frame. In order to achieve high fidelity and generate motions unseen in the input data, VIDIM uses cascaded diffusion models to first generate the target video at low resolution, and then generate the high-resolution video conditioned on the low-resolution generated video. We compare VIDIM to previous state-of-the-art methods on video interpolation, and demonstrate how such works fail in most settings where the underlying motion is complex, nonlinear, or ambiguous while VIDIM can easily handle such cases. We additionally demonstrate how classifier-free guidance on the start and end frame and conditioning the super-resolution model on the original high-resolution frames without additional parameters unlocks high-fidelity results. VIDIM is fast to sample from as it jointly denoises all the frames to be generated, requires less than a billion parameters per diffusion model to produce compelling results, and still enjoys scalability and improved quality at larger parameter counts., Comment: CVPR 2024, Project page at https://vidim-interpolation.github.io/

Generative models are invaluable in many fields of science because of their ability to capture high-dimensional and complicated distributions, such as photo-realistic images, protein structures, and connectomes. How do we evaluate the samples these models generate? This work aims to provide an accessible entry point to understanding popular sample-based statistical distances, requiring only foundational knowledge in mathematics and statistics. We focus on four commonly used notions of statistical distances representing different methodologies: Using low-dimensional projections (Sliced-Wasserstein; SW), obtaining a distance using classifiers (Classifier Two-Sample Tests; C2ST), using embeddings through kernels (Maximum Mean Discrepancy; MMD), or neural networks (Fr\'echet Inception Distance; FID). We highlight the intuition behind each distance and explain their merits, scalability, complexity, and pitfalls. To demonstrate how these distances are used in practice, we evaluate generative models from different scientific domains, namely a model of decision-making and a model generating medical images. We showcase that distinct distances can give different results on similar data. Through this guide, we aim to help researchers to use, interpret, and evaluate statistical distances for generative models in science.

This paper introduces GS-Pose, a unified framework for localizing and estimating the 6D pose of novel objects. GS-Pose begins with a set of posed RGB images of a previously unseen object and builds three distinct representations stored in a database. At inference, GS-Pose operates sequentially by locating the object in the input image, estimating its initial 6D pose using a retrieval approach, and refining the pose with a render-and-compare method. The key insight is the application of the appropriate object representation at each stage of the process. In particular, for the refinement step, we leverage 3D Gaussian splatting, a novel differentiable rendering technique that offers high rendering speed and relatively low optimization time. Off-the-shelf toolchains and commodity hardware, such as mobile phones, can be used to capture new objects to be added to the database. Extensive evaluations on the LINEMOD and OnePose-LowTexture datasets demonstrate excellent performance, establishing the new state-of-the-art. Project page: https://dingdingcai.github.io/gs-pose., Comment: Project Page: https://dingdingcai.github.io/gs-pose

People increasingly rely on online sources for health information seeking due to their convenience and timeliness, traditionally using search engines like Google as the primary search agent. Recently, the emergence of generative Artificial Intelligence (AI) has made Large Language Model (LLM) powered conversational agents such as ChatGPT a viable alternative for health information search. However, while trust is crucial for adopting the online health advice, the factors influencing people's trust judgments in health information provided by LLM-powered conversational agents remain unclear. To address this, we conducted a mixed-methods, within-subjects lab study (N=21) to explore how interactions with different agents (ChatGPT vs. Google) across three health search tasks influence participants' trust judgments of the search results as well as the search agents themselves. Our key findings showed that: (a) participants' trust levels in ChatGPT were significantly higher than Google in the context of health information seeking; (b) there is a significant correlation between trust in health-related information and trust in the search agent, however only for Google; (c) the type of search tasks did not affect participants' perceived trust; and (d) participants' prior knowledge, the style of information presentation, and the interactive manner of using search agents were key determinants of trust in the health-related information. Our study taps into differences in trust perceptions when using traditional search engines compared to LLM-powered conversational agents. We highlight the potential role LLMs play in health-related information-seeking contexts, where they excel as stepping stones for further search. We contribute key factors and considerations for ensuring effective and reliable personal health information seeking in the age of generative AI., Comment: 24 pages

The sizes of subsets of the natural numbers are typically quantified in terms of asymptotic (linear) and logarithmic densities. These concepts have been generalized to weighted $w$-densities, where a specific weight function $w$ plays a key role. In this paper, a parallel theory of asymptotic $\psi$-densities is introduced, where the weight is expressed slightly differently in terms of differentiable functions $\psi$, which are either concave or convex and satisfy certain asymptotic properties. Alternative new proofs for known results on analytic and Abel densities are also given., Comment: 27 pages