Search

The Axion Dark Matter eXperiment is sensitive to narrow axion flows, given axions compose a fraction of the dark matter with a non-negligible local density. Detecting these low-velocity dispersion flows requires a high spectral resolution and careful attention to the expected signal modulation due to Earth's motion. We report an exclusion on the local axion dark matter density in narrow flows of $\rho_a \gtrsim 0.03\,\mathrm{GeV/cm^3}$ and $\rho_a \gtrsim 0.004\,\mathrm{GeV/cm^3}$ for Dine-Fischler-Srednicki-Zhitnitski and Kim-Shifman-Vainshtein-Zakharov axion-photon couplings, respectively, over the mass range $3.3-4.2\,\mu\text{eV}$. Measurements were made at selected resolving powers to allow for a range of possible velocity dispersions., Comment: 7 pages, 3 figures

The 'l-Doubling' phenomenon emanates from the coupling between molecular rotations and perpendicular vibrations (bending modes) in polyatomic molecules. This elusive phenomenon has been largely discarded in laser-induced molecular alignment. Here we explore and unveil the ramifications of 'l-Doubling' to the coherent rotational dynamics of triatomic molecules at ambient temperatures and above. The observed 'l-Doubling' dynamics may be wrongly considered as collisional decay throughout the first few hundreds of picoseconds past excitation, highlighting the importance of correct assimilation of l-Doubling in current research of dissipative rotational dynamics and in coherent rotational dynamics in general., Comment: 9 pages, including supplementary information

One of the most challenging problems in the computational study of localization in quantum manybody systems is to capture the effects of rare events, which requires sampling over exponentially many disorder realizations. We implement an efficient procedure on a quantum processor, leveraging quantum parallelism, to efficiently sample over all disorder realizations. We observe localization without disorder in quantum many-body dynamics in one and two dimensions: perturbations do not diffuse even though both the generator of evolution and the initial states are fully translationally invariant. The disorder strength as well as its density can be readily tuned using the initial state. Furthermore, we demonstrate the versatility of our platform by measuring Renyi entropies. Our method could also be extended to higher moments of the physical observables and disorder learning.

Motivated by a problem in population genetics, we examine the combinatorics of dissimilarity for pairs of random unordered draws of multiple objects, with replacement, from a collection of distinct objects. Consider two draws of size $K$ taken with replacement from a set of $I$ objects, where the two draws represent samples from potentially distinct probability distributions over the set of $I$ objects. We define the set of \emph{identity states} for pairs of draws via a series of actions by permutation groups, describing the enumeration of all such states for a given $K \geq 2$ and $I \geq 2$. Given two probability vectors for the $I$ objects, we compute the probability of each identity state. From the set of all such probabilities, we obtain the expectation for a dissimilarity measure, finding that it has a simple form that generalizes a result previously obtained for the case of $K=2$. We determine when the expected dissimilarity between two draws from the same probability distribution exceeds that of two draws taken from different probability distributions. We interpret the results in the setting of the genetics of polyploid organisms, those whose genetic material contains many copies of the genome ($K > 2$)., Comment: 14 pages, 0 figures

The PlasmaKristall-4 experiment on the International Space Station allows for the study of the 3-dimensional interaction between plasma and dust particles. Previous simulations of the PK-4 environment have discovered fast moving ionization waves in the dc discharge[1]. These ionization waves vary the plasma parameters by up to an order of magnitude, which may affect the mechanisms responsible for the self-organization of chains seen in the PK-4 experiment. Here, we adapt a molecular dynamics simulation to employ temporally varying plasma conditions in order to investigate the effect on the dust charging and electrostatic potential. In order to describe the differences between the average of the plasma conditions and the time-varying plasma condition, we present a model to reproduce the potential that takes into account the negative potential from the dust grain and the positive potential from the ion wake., Comment: The following article has been submitted to Physics of Plasmas

The Colijn--Plazzotta ranking is a bijective encoding of the unlabeled binary rooted trees with positive integers. We show that the rank $f(t)$ of a tree $t$ is closely related to its height $h$, the length of the longest path from a leaf to the root. We consider the rank $f(\tau_n)$ of a random $n$-leaf tree $\tau_n$ under each of three models: (i) uniformly random unlabeled unordered binary rooted trees, or unlabeled topologies; (ii) uniformly random leaf-labeled binary trees, or labeled topologies under the uniform model; and (iii) random binary search trees, or labeled topologies under the Yule--Harding model. Relying on the close relationship between tree rank and tree height, we obtain results concerning the asymptotic properties of $\log \log f(\tau_n)$. In particular, we find $\mathbb{E} \{\log_2 \log f(\tau_n)\} \sim 2 \sqrt{\pi n}$ for uniformly random unlabeled ordered binary rooted trees and uniformly random leaf-labeled binary trees, and for a constant $\alpha \approx 4.31107$, $\mathbb{E}\{\log_2 \log f(\tau_n)\} \sim \alpha \log n $ for leaf-labeled binary trees under the Yule--Harding model. We show that the mean of $f(\tau_n)$ itself under the three models is largely determined by the rank $c_{n-1}$ of the highest-ranked tree -- the caterpillar -- obtaining an asymptotic relationship with $\pi_n c_{n-1}$, where $\pi_n$ is a model-specific function of $n$. The results resolve open problems, providing a new class of results on an encoding useful in mathematical phylogenetics.

We study the cohomology rings of tiling spaces $\Omega$ given by cubical substitutions. While there have been many calculations before of cohomology groups of such tiling spaces, the innovation here is that we use computer-assisted methods to compute the cup-product structure. This leads to examples of substitution tilings with isomorphic cohomology groups but different cohomology rings. Part of the interest in studying the cup product comes from Bellissard's gap-labeling conjecture, which is known to hold in dimensions $\le 3$, but where a proof is known in dimensions $\ge 4$ only when the Chern character from $K^0(\Omega)$ to $H^*(\Omega,\mathbb{Q})$ lands in $H^*(\Omega,\mathbb{Z})$. Computation of the cup product on cohomology often makes it possible to compute the Chern character. We introduce a natural generalization of the gap-labeling conjecture, called the equivariant gap-labeling conjecture, which applies to tilings with a finite symmetry group. Again this holds in dimensions $\le 3$, but we are able to show that it fails in general in dimensions $\ge 4$. This, plus some of our cup product calculations, makes it plausible that the gap-labeling conjecture might fail in high dimensions., Comment: Fixed mistake in the first pair of examples, added references

Lattice gauge theories (LGTs) can be employed to understand a wide range of phenomena, from elementary particle scattering in high-energy physics to effective descriptions of many-body interactions in materials. Studying dynamical properties of emergent phases can be challenging as it requires solving many-body problems that are generally beyond perturbative limits. We investigate the dynamics of local excitations in a $\mathbb{Z}_2$ LGT using a two-dimensional lattice of superconducting qubits. We first construct a simple variational circuit which prepares low-energy states that have a large overlap with the ground state; then we create particles with local gates and simulate their quantum dynamics via a discretized time evolution. As the effective magnetic field is increased, our measurements show signatures of transitioning from deconfined to confined dynamics. For confined excitations, the magnetic field induces a tension in the string connecting them. Our method allows us to experimentally image string dynamics in a (2+1)D LGT from which we uncover two distinct regimes inside the confining phase: for weak confinement the string fluctuates strongly in the transverse direction, while for strong confinement transverse fluctuations are effectively frozen. In addition, we demonstrate a resonance condition at which dynamical string breaking is facilitated. Our LGT implementation on a quantum processor presents a novel set of techniques for investigating emergent particle and string dynamics.

This short note summarizes a number of facts about the ring $K^0(X)$ for $X$ a $4$-dimensional CW-complex. Unusual features of this dimension are that every complex vector bundle is determined up to stable isomorphism by its Chern classes, that every even cohomology class arises as a Chern class of a vector bundle, and that $K^0(X)$ is completely determined as a ring by knowledge of the even-dimensional cohomology ring $H^{\text{even}}(X; \mathbb Z)$. (All of these fail in high dimensions.), Comment: 4 pages. some typo corrections and one new reference

Conversational Swarm Intelligence (CSI) is an AI-powered communication and collaboration technology that allows large, networked groups (of potentially unlimited size) to hold thoughtful conversational deliberations in real-time. Inspired by the efficient decision-making dynamics of fish schools, CSI divides a human population into a set of small subgroups connected by AI agents. This enables the full group to hold a unified conversation. In this study, groups of 25 participants were tasked with selecting a roster of players in a real Fantasy Baseball contest. A total of 10 trials were run using CSI. In half the trials, each subgroup was augmented with a fact-providing AI agent referred to herein as an Infobot. The Infobot was loaded with a wide range of MLB statistics. The human participants could query the Infobot the same way they would query other persons in their subgroup. Results show that when using CSI, the 25-person groups outperformed 72% of individually surveyed participants and showed significant intelligence amplification versus the mean score (p=0.016). The CSI-enabled groups also significantly outperformed the most popular picks across the collected surveys for each daily contest (p<0.001). The CSI sessions that used Infobots scored slightly higher than those that did not, but it was not statistically significant in this study. That said, 85% of participants agreed with the statement 'Our decisions were stronger because of information provided by the Infobot' and only 4% disagreed. In addition, deliberations that used Infobots showed significantly less variance (p=0.039) in conversational content across members. This suggests that Infobots promoted more balanced discussions in which fewer members dominated the dialog. This may be because the infobot enabled participants to confidently express opinions with the support of factual data

In this paper, we introduce a zero-shot Voice Transfer (VT) module that can be seamlessly integrated into a multi-lingual Text-to-speech (TTS) system to transfer an individual's voice across languages. Our proposed VT module comprises a speaker-encoder that processes reference speech, a bottleneck layer, and residual adapters, connected to preexisting TTS layers. We compare the performance of various configurations of these components and report Mean Opinion Score (MOS) and Speaker Similarity across languages. Using a single English reference speech per speaker, we achieve an average voice transfer similarity score of 73% across nine target languages. Vocal characteristics contribute significantly to the construction and perception of individual identity. The loss of one's voice, due to physical or neurological conditions, can lead to a profound sense of loss, impacting one's core identity. As a case study, we demonstrate that our approach can not only transfer typical speech but also restore the voices of individuals with dysarthria, even when only atypical speech samples are available - a valuable utility for those who have never had typical speech or banked their voice. Cross-lingual typical audio samples, plus videos demonstrating voice restoration for dysarthric speakers are available here (google.github.io/tacotron/publications/zero_shot_voice_transfer)., Comment: Submitted to ICASSP

We explore the possibility to implement random walks in the manifold of Hartree-Fock-Bogoliubov wave functions. The goal is to extend state-of-the-art quantum Monte Carlo approaches, in particular the constrained-path auxiliary-field quantum Monte Carlo technique, to systems where finite pairing order parameters or complex pairing mechanisms, e.g., Fulde-Ferrell-Larkin-Ovchinnikov (FFLO) pairing or triplet pairing, may be expected. Leveraging the flexibility to define a vacuum state tailored to the physical problem, we discuss a method to use imaginary-time evolution of Hartree-Fock-Bogoliubov states to compute ground state correlations, extending beyond situations spanned by current formalisms. Illustrative examples are provided., Comment: 11 pages, 3 figures

Bilevel optimization methods are increasingly relevant within machine learning, especially for tasks such as hyperparameter optimization and meta-learning. Compared to the offline setting, online bilevel optimization (OBO) offers a more dynamic framework by accommodating time-varying functions and sequentially arriving data. This study addresses the online nonconvex-strongly convex bilevel optimization problem. In deterministic settings, we introduce a novel online Bregman bilevel optimizer (OBBO) that utilizes adaptive Bregman divergences. We demonstrate that OBBO enhances the known sublinear rates for bilevel local regret through a novel hypergradient error decomposition that adapts to the underlying geometry of the problem. In stochastic contexts, we introduce the first stochastic online bilevel optimizer (SOBBO), which employs a window averaging method for updating outer-level variables using a weighted average of recent stochastic approximations of hypergradients. This approach not only achieves sublinear rates of bilevel local regret but also serves as an effective variance reduction strategy, obviating the need for additional stochastic gradient samples at each timestep. Experiments on online hyperparameter optimization and online meta-learning highlight the superior performance, efficiency, and adaptability of our Bregman-based algorithms compared to established online and offline bilevel benchmarks.

Laser-plasma instabilities under ignition conditions for direct-drive inertial confinement fusion are studied using two-dimensional Particle-in-Cell simulations with a combination of in-plane (PP) and out-of-the-plane (SP) lasers. The results show that stimulated Raman side scattering can induce significant pump depletion and form a gap in the Raman scattered light spectra that have been observed in experiments., Comment: 6 pages, 4 figures

Recent studies have shown that large language models' (LLMs) mathematical problem-solving capabilities can be enhanced by integrating external tools, such as code interpreters, and employing multi-turn Chain-of-Thought (CoT) reasoning. While current methods focus on synthetic data generation and Supervised Fine-Tuning (SFT), this paper studies the complementary direct preference learning approach to further improve model performance. However, existing direct preference learning algorithms are originally designed for the single-turn chat task, and do not fully address the complexities of multi-turn reasoning and external tool integration required for tool-integrated mathematical reasoning tasks. To fill in this gap, we introduce a multi-turn direct preference learning framework, tailored for this context, that leverages feedback from code interpreters and optimizes trajectory-level preferences. This framework includes multi-turn DPO and multi-turn KTO as specific implementations. The effectiveness of our framework is validated through training of various language models using an augmented prompt set from the GSM8K and MATH datasets. Our results demonstrate substantial improvements: a supervised fine-tuned Gemma-1.1-it-7B model's performance increased from 77.5% to 83.9% on GSM8K and from 46.1% to 51.2% on MATH. Similarly, a Gemma-2-it-9B model improved from 84.1% to 86.3% on GSM8K and from 51.0% to 54.5% on MATH., Comment: A multi-turn direct preference learning framework for tool-integrated reasoning tasks

Representing speech as discrete tokens provides a framework for transforming speech into a format that closely resembles text, thus enabling the use of speech as an input to the widely successful large language models (LLMs). Currently, while several speech tokenizers have been proposed, there is ambiguity regarding the properties that are desired from a tokenizer for specific downstream tasks and its overall generalizability. Evaluating the performance of tokenizers across different downstream tasks is a computationally intensive effort that poses challenges for scalability. To circumvent this requirement, we present STAB (Speech Tokenizer Assessment Benchmark), a systematic evaluation framework designed to assess speech tokenizers comprehensively and shed light on their inherent characteristics. This framework provides a deeper understanding of the underlying mechanisms of speech tokenization, thereby offering a valuable resource for expediting the advancement of future tokenizer models and enabling comparative analysis using a standardized benchmark. We evaluate the STAB metrics and correlate this with downstream task performance across a range of speech tasks and tokenizer choices., Comment: 5 pages

We report the results of a QCD axion dark matter search with discovery ability for Dine-Fischler-Srednicki-Zhitnitsky (DFSZ) axions using an axion haloscope. Sub-Kelvin noise temperatures are reached with an ultra low-noise Josephson parametric amplifier cooled by a dilution refrigerator. This work excludes (with a 90% confidence level) DFSZ axions with masses between 3.27 to 3.34 {\mu}eV, assuming a standard halo model with a local energy density of 0.45 GeV/cc made up 100% of axions.

Quantum error correction provides a path to reach practical quantum computing by combining multiple physical qubits into a logical qubit, where the logical error rate is suppressed exponentially as more qubits are added. However, this exponential suppression only occurs if the physical error rate is below a critical threshold. In this work, we present two surface code memories operating below this threshold: a distance-7 code and a distance-5 code integrated with a real-time decoder. The logical error rate of our larger quantum memory is suppressed by a factor of $\Lambda$ = 2.14 $\pm$ 0.02 when increasing the code distance by two, culminating in a 101-qubit distance-7 code with 0.143% $\pm$ 0.003% error per cycle of error correction. This logical memory is also beyond break-even, exceeding its best physical qubit's lifetime by a factor of 2.4 $\pm$ 0.3. We maintain below-threshold performance when decoding in real time, achieving an average decoder latency of 63 $\mu$s at distance-5 up to a million cycles, with a cycle time of 1.1 $\mu$s. To probe the limits of our error-correction performance, we run repetition codes up to distance-29 and find that logical performance is limited by rare correlated error events occurring approximately once every hour, or 3 $\times$ 10$^9$ cycles. Our results present device performance that, if scaled, could realize the operational requirements of large scale fault-tolerant quantum algorithms., Comment: 10 pages, 4 figures, Supplementary Information

Proton radiography is a central diagnostic technique for measuring electromagnetic (EM) fields in high-energy-density, laser-produced plasmas. In this technique, protons traverse the plasma where they accumulate small EM deflections which lead to variations in the proton fluence pattern on a detector. Path-integrated EM fields can then be extracted from the fluence image through an inversion process. In this work, experiments of laser-driven foils were conducted on the OMEGA laser and magnetic field reconstructions were performed using both "fluence-based" techniques and high-fidelity "mesh-based" methods. We implement nonzero boundary conditions into the inversion and show their importance by comparing against mesh measurements. Good agreement between the methods is found only when nonzero boundary conditions are used. We also introduce an approach to determine the unperturbed proton source profile, which is a required input in fluence reconstruction algorithms. In this approach, a fluence inversion is embedded inside of a mesh region, which provides overconstrained magnetic boundary conditions. A source profile is then iteratively optimized to satisfy the boundary information. This method substantially enhances the accuracy in recovering EM fields. Lastly, we propose a scheme to quantify uncertainty in the final inversion that is introduced through errors in the source retrieval., Comment: 16 pages, 12 figures

The keyword spotting (KWS) problem requires large amounts of real speech training data to achieve high accuracy across diverse populations. Utilizing large amounts of text-to-speech (TTS) synthesized data can reduce the cost and time associated with KWS development. However, TTS data may contain artifacts not present in real speech, which the KWS model can exploit (overfit), leading to degraded accuracy on real speech. To address this issue, we propose applying an adversarial training method to prevent the KWS model from learning TTS-specific features when trained on large amounts of TTS data. Experimental results demonstrate that KWS model accuracy on real speech data can be improved by up to 12% when adversarial loss is used in addition to the original KWS loss. Surprisingly, we also observed that the adversarial setup improves accuracy by up to 8%, even when trained solely on TTS and real negative speech data, without any real positive examples., Comment: to be published in a Workshop at Interspeech 2024, Synthetic Data's Transformative Role in Foundational Speech Models

This paper explores the use of TTS synthesized training data for KWS (keyword spotting) task while minimizing development cost and time. Keyword spotting models require a huge amount of training data to be accurate, and obtaining such training data can be costly. In the current state of the art, TTS models can generate large amounts of natural-sounding data, which can help reducing cost and time for KWS model development. Still, TTS generated data can be lacking diversity compared to real data. To pursue maximizing KWS model accuracy under the constraint of limited resources and current TTS capability, we explored various strategies to mix TTS data and real human speech data, with a focus on minimizing real data use and maximizing diversity of TTS output. Our experimental results indicate that relatively small amounts of real audio data with speaker diversity (100 speakers, 2k utterances) and large amounts of TTS synthesized data can achieve reasonably high accuracy (within 3x error rate of baseline), compared to the baseline (trained with 3.8M real positive utterances)., Comment: to be published in a Workshop at Interspeech 2024, Synthetic Data's Transformative Role in Foundational Speech Models

Doping impurity atoms is a strategy commonly used to tune the functionality of materials including catalysts, semiconductors, and quantum emitters. The location of dopants and their interaction with surrounding atoms could significantly modulate the transport, optical, or magnetic properties of materials. However, directly imaging individual impurity atoms inside materials remains a generally unaddressed need. Here, we demonstrate how single atoms can be detected and located in three dimensions via multislice electron ptychography.Interstitial atoms in a complex garnet oxide heterostructure are resolved with a depth resolution better than 2.7 nm, together with a deep-sub-{\AA}ngstrom lateral resolution. Single-scan atomic-layer depth resolution should be possible using strongly divergent electron probe illumination. Our results provide a new approach to detecting individual atomic defects and open doors to characterize the local environments and spatial distributions that underlie a broad range of systems such as single-atom catalysts, nitrogen-vacancy centers, and other atomic-scale quantum sensors., Comment: 28 pages, 5 figures, 10 supplementary figures

Computer vision systems have been deployed in various applications involving biometrics like human faces. These systems can identify social media users, search for missing persons, and verify identity of individuals. While computer vision models are often evaluated for accuracy on available benchmarks, more annotated data is necessary to learn about their robustness and fairness against semantic distributional shifts in input data, especially in face data. Among annotated data, counterfactual examples grant strong explainability characteristics. Because collecting natural face data is prohibitively expensive, we put forth a generative AI-based framework to construct targeted, counterfactual, high-quality synthetic face data. Our synthetic data pipeline has many use cases, including face recognition systems sensitivity evaluations and image understanding system probes. The pipeline is validated with multiple user studies. We showcase the efficacy of our face generation pipeline on a leading commercial vision model. We identify facial attributes that cause vision systems to fail., Comment: Paper under review. Full text and results will be updated after acceptance

This paper reports on a measurement of electron-ion recombination in liquid argon in the ICARUS liquid argon time projection chamber (LArTPC). A clear dependence of recombination on the angle of the ionizing particle track relative to the drift electric field is observed. An ellipsoid modified box (EMB) model of recombination describes the data across all measured angles. These measurements are used for the calorimetric energy scale calibration of the ICARUS TPC, which is also presented. The impact of the EMB model is studied on calorimetric particle identification, as well as muon and proton energy measurements. Accounting for the angular dependence in EMB recombination improves the accuracy and precision of these measurements.

The ICARUS liquid argon time projection chamber (LArTPC) neutrino detector has been taking physics data since 2022 as part of the Short-Baseline Neutrino (SBN) Program. This paper details the equalization of the response to charge in the ICARUS time projection chamber (TPC), as well as data-driven tuning of the simulation of ionization charge signals and electronics noise. The equalization procedure removes non-uniformities in the ICARUS TPC response to charge in space and time. This work leverages the copious number of cosmic ray muons available to ICARUS at the surface. The ionization signal shape simulation applies a novel procedure that tunes the simulation to match what is measured in data. The end result of the equalization procedure and simulation tuning allows for a comparison of charge measurements in ICARUS between Monte Carlo simulation and data, showing good performance with minimal residual bias between the two.

We perform readout of a quantum-dot hybrid qubit coupled to a superconducting resonator through a parametric, longitudinal interaction mechanism. Our experiments are performed with the qubit and resonator frequencies detuned by $\sim$10 GHz, demonstrating that longitudinal coupling can facilitate semiconductor qubit operation in the 'ultra-dispersive' regime of circuit quantum electrodynamics., Comment: Main text and Supplementary Information; 12 pages, 7 figures

This paper explores speculative speech recognition (SSR), where we empower conventional automatic speech recognition (ASR) with speculation capabilities, allowing the recognizer to run ahead of audio. We introduce a metric for measuring SSR performance and we propose a model which does SSR by combining a RNN-Transducer-based ASR system with an audio-prefixed language model (LM). The ASR system transcribes ongoing audio and feeds the resulting transcripts, along with an audio-dependent prefix, to the LM, which speculates likely completions for the transcriptions. We experiment with a variety of ASR datasets on which show the efficacy our method and the feasibility of SSR as a method of reducing ASR latency., Comment: Interspeech 2024

Policy Optimization (PO) methods are among the most popular Reinforcement Learning (RL) algorithms in practice. Recently, Sherman et al. [2023a] proposed a PO-based algorithm with rate-optimal regret guarantees under the linear Markov Decision Process (MDP) model. However, their algorithm relies on a costly pure exploration warm-up phase that is hard to implement in practice. This paper eliminates this undesired warm-up phase, replacing it with a simple and efficient contraction mechanism. Our PO algorithm achieves rate-optimal regret with improved dependence on the other parameters of the problem (horizon and function approximation dimension) in two fundamental settings: adversarial losses with full-information feedback and stochastic losses with bandit feedback.

Open-source software and Commercial Off-The-Shelf hardware are finally paving their way into the 5G world, resulting in a proliferation of experimental 5G testbeds. Surprisingly, very few studies have been published on the comparative analysis of testbeds with different hardware and software elements. In this paper, we first introduce a precise nomenclature to characterize a 5G-standalone single-cell testbed based on its constituent elements and main configuration parameters. We then build 30 distinct such testbeds and systematically analyze their performance with an emphasis on element interoperability (by considering different combinations of hardware and software elements from different sources), the number and type of User Equipment (UE) as well as the Radio Access Network hardware and software elements to address the following questions: 1) How is the performance (in terms of bit rate and latency) impacted by different elements? 2) How does the number of UEs affect these results? 3) What is the impact of the user(s)' location(s) on the performance? 4) What is the impact of the UE type on these results? 5) How far does each testbed provide coverage? 6) And finally, what is the effect of the computing resources available to each open-source software? This study focuses on TDD testbeds.

Multimodal Retrieval Augmented Generation (MMRAG) is a powerful approach to question-answering over multimodal documents. A key challenge with evaluating MMRAG is the paucity of high-quality datasets matching the question styles and modalities of interest. In light of this, we propose SMMQG, a synthetic data generation framework. SMMQG leverages interplay between a retriever, large language model (LLM) and large multimodal model (LMM) to generate question and answer pairs directly from multimodal documents, with the questions conforming to specified styles and modalities. We use SMMQG to generate an MMRAG dataset of 1024 questions over Wikipedia documents and evaluate state-of-the-art models using it, revealing insights into model performance that are attainable only through style- and modality-specific evaluation data. Next, we measure the quality of data produced by SMMQG via a human study. We find that the quality of SMMQG-generated synthetic data is on par with the quality of the crowdsourced benchmark MMQA and that downstream evaluation results using both datasets strongly concur., Comment: Accepted to EMNLP 2024 Findings; Camera Ready

In this paper, we focus on addressing the constraints faced when applying LLMs to ASR. Recent works utilize prefixLM-type models, which directly apply speech as a prefix to LLMs for ASR. We have found that optimizing speech prefixes leads to better ASR performance and propose applying RNNT loss to perform speech prefix-tuning. This is a simple approach and does not increase the model complexity or alter the inference pipeline. We also propose language-based soft prompting to further improve with frozen LLMs. Empirical analysis on realtime testset from 10 Indic languages demonstrate that our proposed speech prefix-tuning yields improvements with both frozen and fine-tuned LLMs. Our recognition results on an average of 10 Indics show that the proposed prefix-tuning with RNNT loss results in a 12\% relative improvement in WER over the baseline with a fine-tuned LLM. Our proposed approches with the frozen LLM leads to a 31\% relative improvement over basic soft-prompting prefixLM.

The quantum mechanical nature of spherical top molecules is particularly evident at low angular momentum quantum number J. Using infrared spectroscopy on the 8.4$\mu$m rovibrational band of buffer gas cooled $^{12}$C$_{60}$, we observe the hitherto unseen R(J = 0 - 29) rotational progression, including the complete disappearance of certain transitions due to the molecule's perfect icosahedral symmetry and identical bosonic nuclei. The observation of extremely weak C$_{60}$ absorption is facilitated by a laser desorption C$_{60}$ vapor source, which transfers 1000-fold less heat to the cryogenic buffer gas cell than a traditional oven source. This technique paves the way to cooling C$_{60}$ and other large gas phase molecules to much lower temperatures, providing continued advances for spectral resolution and sensitivity.

A significant challenge in measurements of neutrino oscillations is reconstructing the incoming neutrino energies. While modern fully-active tracking calorimeters such as liquid argon time projection chambers in principle allow the measurement of all final state particles above some detection threshold, undetected neutrons remain a considerable source of missing energy with little to no data constraining their production rates and kinematics. We present the first demonstration of tagging neutrino-induced neutrons in liquid argon time projection chambers using secondary protons emitted from neutron-argon interactions in the MicroBooNE detector. We describe the method developed to identify neutrino-induced neutrons and demonstrate its performance using neutrons produced in muon-neutrino charged current interactions. The method is validated using a small subset of MicroBooNE's total dataset. The selection yields a sample with $60\%$ of selected tracks corresponding to neutron-induced secondary protons.

We present a deep learning-based method for estimating the neutrino energy of charged-current neutrino-argon interactions. We employ a recurrent neural network (RNN) architecture for neutrino energy estimation in the MicroBooNE experiment, utilizing liquid argon time projection chamber (LArTPC) detector technology. Traditional energy estimation approaches in LArTPCs, which largely rely on reconstructing and summing visible energies, often experience sizable biases and resolution smearing because of the complex nature of neutrino interactions and the detector response. The estimation of neutrino energy can be improved after considering the kinematics information of reconstructed final-state particles. Utilizing kinematic information of reconstructed particles, the deep learning-based approach shows improved resolution and reduced bias for the muon neutrino Monte Carlo simulation sample compared to the traditional approach. In order to address the common concern about the effectiveness of this method on experimental data, the RNN-based energy estimator is further examined and validated with dedicated data-simulation consistency tests using MicroBooNE data. We also assess its potential impact on a neutrino oscillation study after accounting for all statistical and systematic uncertainties and show that it enhances physics sensitivity. This method has good potential to improve the performance of other physics analyses.

This paper introduces ASTRA, a novel method for improving Automatic Speech Recognition (ASR) through text injection.Unlike prevailing techniques, ASTRA eliminates the need for sampling to match sequence lengths between speech and text modalities. Instead, it leverages the inherent alignments learned within CTC/RNNT models. This approach offers the following two advantages, namely, avoiding potential misalignment between speech and text features that could arise from upsampling and eliminating the need for models to accurately predict duration of sub-word tokens. This novel formulation of modality (length) matching as a weighted RNNT objective matches the performance of the state-of-the-art duration-based methods on the FLEURS benchmark, while opening up other avenues of research in speech processing., Comment: To be published in Interspeech 2024

Understanding how interacting particles approach thermal equilibrium is a major challenge of quantum simulators. Unlocking the full potential of such systems toward this goal requires flexible initial state preparation, precise time evolution, and extensive probes for final state characterization. We present a quantum simulator comprising 69 superconducting qubits which supports both universal quantum gates and high-fidelity analog evolution, with performance beyond the reach of classical simulation in cross-entropy benchmarking experiments. Emulating a two-dimensional (2D) XY quantum magnet, we leverage a wide range of measurement techniques to study quantum states after ramps from an antiferromagnetic initial state. We observe signatures of the classical Kosterlitz-Thouless phase transition, as well as strong deviations from Kibble-Zurek scaling predictions attributed to the interplay between quantum and classical coarsening of the correlated domains. This interpretation is corroborated by injecting variable energy density into the initial state, which enables studying the effects of the eigenstate thermalization hypothesis (ETH) in targeted parts of the eigenspectrum. Finally, we digitally prepare the system in pairwise-entangled dimer states and image the transport of energy and vorticity during thermalization. These results establish the efficacy of superconducting analog-digital quantum processors for preparing states across many-body spectra and unveiling their thermalization dynamics.

Reinforcement Learning from Human Feedback (RLHF) has become the standard approach for aligning Large Language Models (LLMs) with human preferences, allowing LLMs to demonstrate remarkable abilities in various tasks. Existing methods work by emulating the preferences at the single decision (turn) level, limiting their capabilities in settings that require planning or multi-turn interactions to achieve a long-term goal. In this paper, we address this issue by developing novel methods for Reinforcement Learning (RL) from preference feedback between two full multi-turn conversations. In the tabular setting, we present a novel mirror-descent-based policy optimization algorithm for the general multi-turn preference-based RL problem, and prove its convergence to Nash equilibrium. To evaluate performance, we create a new environment, Education Dialogue, where a teacher agent guides a student in learning a random topic, and show that a deep RL variant of our algorithm outperforms RLHF baselines. Finally, we show that in an environment with explicit rewards, our algorithm recovers the same performance as a reward-based RL baseline, despite relying solely on a weaker preference signal.

In many real-world applications, it is hard to provide a reward signal in each step of a Reinforcement Learning (RL) process and more natural to give feedback when an episode ends. To this end, we study the recently proposed model of RL with Aggregate Bandit Feedback (RL-ABF), where the agent only observes the sum of rewards at the end of an episode instead of each reward individually. Prior work studied RL-ABF only in tabular settings, where the number of states is assumed to be small. In this paper, we extend ABF to linear function approximation and develop two efficient algorithms with near-optimal regret guarantees: a value-based optimistic algorithm built on a new randomization technique with a Q-functions ensemble, and a policy optimization algorithm that uses a novel hedging scheme over the ensemble.

This paper contains corrections to Madea, Rosenberg, Torres-Ardila, "The Geometry of Loop Spaces II: Characteristic Classes," Advances in Math. (287), 2016, 485-518. The main change is that results about $\pi_1({\rm Diff}(M))$ are replaced by results about $\pi_1({\rm Isom}(M))$, where Diff$(M)$, Isom$(M)$ refer to the diffeomorphism and isometry group of the manifold $M$.

Conducting rigorous impact studies that produce actionable evidence for educators and policymakers is a key goal of the Standards for Excellence in Education Research (SEER). Although studies can be initiated by individual districts or schools, researchers often conduct their own studies to address broadly relevant questions about the effectiveness of an education program, policy, practice, or intervention. This brief offers practical strategies for researchers newer to school-based impact studies to consider during the various phases of a study, with the aim of helping them achieve their recruitment goals. The strategies focus on four topics that, based in part on the authors' knowledge of recent recruitment efforts, may play an increasingly important role in affecting a district's or school's ability and willingness to participate in impact studies: (1) Attention to urgent priorities and needs; (2) Limited staff capacity; (3) Commitment to fairness and collaborative practices; and (4) Protection of student and staff privacy.