Negative / Null Result ReportOpen accessComputer Science
Chenming Tang, Xiuyu Wu, Yunfang Wu · 2023 · arXiv
Model ensemble has been in widespread use for Grammatical Error Correction (GEC), boosting model performance. We hypothesize that model ensemble based on the perplexity (PPL) computed by pre-trained language models (PLMs) should benefit the GEC system. To this end, we explore several ensemble strategies based on strong PLMs with four sophisticated single models. However, the performance does not improve but even gets worse after the PLM-based ensemble. This surprising result sets us doing a detailed analysis on the data and coming up with some insights on GEC. The human references of correct s
Negative / Null Result ReportOpen accessComputer Science
Guanzi Chen, Jiying Zhang, Xi Xiao et al. · 2022 · arXiv
In recent years, hypergraph learning has attracted great attention due to its capacity in representing complex and high-order relationships. However, current neural network approaches designed for hypergraphs are mostly shallow, thus limiting their ability to extract information from high-order neighbors. In this paper, we show both theoretically and empirically, that the performance of hypergraph neural networks does not improve as the number of layers increases, which is known as the over-smoothing problem. To avoid this issue, we develop a new deep hypergraph convolutional network called De
Negative / Null Result ReportOpen accessEngineering
Siqi Ye, Yong Long, Il Yong Chun · 2020 · arXiv
This paper applies the recent fast iterative neural network framework, Momentum-Net, using appropriate models to low-dose X-ray computed tomography (LDCT) image reconstruction. At each layer of the proposed Momentum-Net, the model-based image reconstruction module solves the majorized penalized weighted least-square problem, and the image refining module uses a four-layer convolutional neural network (CNN). Experimental results with the NIH AAPM-Mayo Clinic Low Dose CT Grand Challenge dataset show that the proposed Momentum-Net architecture significantly improves image reconstruction accuracy,
Negative / Null Result ReportOpen accessComputer Science
Daniel Mann, Tina Raissi, Wilfried Michel et al. · 2023 · arXiv
We investigate a novel modeling approach for end-to-end neural network training using hidden Markov models (HMM) where the transition probabilities between hidden states are modeled and learned explicitly. Most contemporary sequence-to-sequence models allow for from-scratch training by summing over all possible label segmentations in a given topology. In our approach there are explicit, learnable probabilities for transitions between segments as opposed to a blank label that implicitly encodes duration statistics. We implement a GPU-based forward-backward algorithm that enables the simultaneou
Negative / Null Result ReportOpen accessAgricultural and Biological Sciences
Roy Henha Eyono, Daniel Levenstein, Arna Ghosh et al. · 2026 · arXiv
Normalization is a critical operation in neural circuits. In the brain, there is evidence that normalization is implemented via inhibitory interneurons and allows neural populations to adjust to changes in the distribution of their inputs. In artificial neural networks (ANNs), normalization is used to improve learning in tasks that involve complex input distributions. However, it is unclear whether inhibition-mediated normalization in biological neural circuits also improves learning. Here, we explore this possibility using ANNs with separate excitatory and inhibitory populations trained on an
Negative / Null Result ReportOpen accessPhysics
Susanne Still, David A. Sivak, Anthony J. Bell et al. · 2012 · arXiv
A system responding to a stochastic driving signal can be interpreted as computing, by means of its dynamics, an implicit model of the environmental variables. The system's state retains information about past environmental fluctuations, and a fraction of this information is predictive of future ones. The remaining nonpredictive information reflects model complexity that does not improve predictive power, and thus represents the ineffectiveness of the model. We expose the fundamental equivalence between this model inefficiency and thermodynamic inefficiency, measured by dissipation. Our result
Negative / Null Result ReportOpen accessComputer Science
M. M. Sargsian, S. Simula, M. I. Strikman · 2002 · arXiv
A detailed study of inclusive deep inelastic scattering from mirror A = 3 nuclei at large values of Bjorken-x is presented. The main purpose is to estimate the theoretical uncertainties on the extraction of F2n from such measurements. Within the convolution approach we confirm the cancellation of nuclear effects at the level of ~1 % for x < 0.75 in overall agreement with previous findings. However, within models in which modifications of the bound nucleon structure functions are accounted for to describe the EMC effect in nuclei, we find that the nuclear effects may be canceled at a level of ~
Negative / Null Result ReportOpen accessComputer Science
Solène Tarride, Tristan Faine, Mélodie Boillet et al. · 2023 · arXiv
In this paper, we explore different ways of training a model for handwritten text recognition when multiple imperfect or noisy transcriptions are available. We consider various training configurations, such as selecting a single transcription, retaining all transcriptions, or computing an aggregated transcription from all available annotations. In addition, we evaluate the impact of quality-based data selection, where samples with low agreement are removed from the training set. Our experiments are carried out on municipal registers of the city of Belfort (France) written between 1790 and 1946
Negative / Null Result ReportOpen accessPhysics
Ilias Magoulas, Karthik Gururangan, Piotr Piecuch et al. · 2021 · arXiv
The short answer to the question in the title is 'no'. We identify classes of truncated configuration interaction (CI) wave functions for which the externally corrected coupled-cluster (ec-CC) approach using the three-body ($T_{3}$) and four-body ($T_{4}$) components of the cluster operator extracted from CI does not improve the results of the underlying CI calculations. Implications of our analysis, illustrated by numerical examples, for the ec-CC computations using truncated and selected CI methods are discussed. We also introduce a novel ec-CC approach using the $T_{3}$ and $T_{4}$ amplitud
Negative / Null Result ReportOpen accessComputer Science
Syomantak Chaudhuri, Thomas A. Courtade · 2023 · arXiv
Differential Privacy (DP) is a well-established framework to quantify privacy loss incurred by any algorithm. Traditional DP formulations impose a uniform privacy requirement for all users, which is often inconsistent with real-world scenarios in which users dictate their privacy preferences individually. This work considers the problem of mean estimation under heterogeneous DP constraints, where each user can impose their own distinct privacy level. The algorithm we propose is shown to be minimax optimal when there are two groups of users with distinct privacy levels. Our results elicit an in
Negative / Null Result ReportOpen accessComputer Science
T. C. Liu, M. A. Huang, Guey-Lin Lin · 2010 · arXiv
We discuss the reconstruction of neutrino flavor neutrino at a distant source in the very high en- ergy regime. This reconstruction procedure is relevant to the confirmation of detecting cosmogenic neutrinos, for example. To facilitate such a reconstruction, it is imperative to achieve effective flavor discriminations in terrestrial neutrino telescopes. We note that, for energies beyond few tens of PeV, a tau-lepton behaves like a track similar to a muon. Hence, while it is rather challenging to separate νμ from ντ in this case, one can expect to isolate νe from the rest by a distinctive showe
Negative / Null Result ReportOpen accessComputer Science
Seyed Arash Ghoreishi, Nidhin Sudarsanan Ragini, Sk Sazim et al. · 2024 · arXiv
Quantum labeling tasks ask one to recover the missing associations between classical outcome labels and the effects forming the POVM. We study labeling in the multiple-shot regime, allowing a finite number of uses of the device and the most general tester-based strategies, including adaptivity. For binary observables, we show that if perfect labeling is impossible in a single shot, then it remains impossible with any finite number of shots. In particular, we derive the formula for minimum-error performance and highlight its ``even-odd" behavior. For non-binary observables, we derive the optima
Negative / Null Result ReportOpen accessPhysics
Lehman H. Garrison, Michael Joyce, Daniel J. Eisenstein · 2021 · arXiv
Analysis of self-similarity in scale-free $N$-body simulations reveals the spatial and temporal scales for which statistics measured in cosmological simulations are converged to the physical continuum limit. We examine how the range of scales in which the two-point correlation function is converged depends on the force softening length and whether it is held constant in comoving or proper coordinates. We find that a proper softening that reaches roughly 1/30th of the inter-particle spacing by the end of the simulation resolves the same spatial and temporal scales as a comoving softening of the
Negative / Null Result ReportOpen accessComputer Science
Michele Filannino, Gavin Brown, Goran Nenadic · 2013 · arXiv
This paper describes a temporal expression identification and normalization system, ManTIME, developed for the TempEval-3 challenge. The identification phase combines the use of conditional random fields along with a post-processing identification pipeline, whereas the normalization phase is carried out using NorMA, an open-source rule-based temporal normalizer. We investigate the performance variation with respect to different feature types. Specifically, we show that the use of WordNet-based features in the identification task negatively affects the overall performance, and that there is no
Negative / Null Result ReportOpen accessMedicine
Mayer L, Lichtblau M, Furian M et al. · 2025 · Journal of the American Heart Association
Background Exposure to high altitude triggers hypoxic pulmonary vasoconstriction with a rise in pulmonary artery pressure (PAP). This may lead to symptomatic pulmonary hypertension with impaired right heart function and exercise capacity.…
View details →DOI: 10.1161/jaha.125.041941 Negative / Null Result ReportOpen accessMedicine
López-de la Rosa A, Díez-Montero C, Martínez-Plaza E et al. · 2024 · Ophthalmology and therapy
Introduction Characterizing lens thickness (LT) in patients with cataracts is important for better understanding the lens aging process and for designing new intraocular lens power formulas. This study aimed to analyze the influence of…
View details →DOI: 10.1007/s40123-024-00882-6 Negative / Null Result ReportMedicine
Fitzgerald LS, Reynoso Spurrier CS, Lau NJ et al. · 2025 · American journal of physiology. Endocrinology and metabolism
Mitochondria are key regulators of metabolism and ATP supply in skeletal muscle, while circadian rhythms influence many physiological processes. However, whether mitochondrial function is intrinsically regulated in a circadian manner in…
View details →DOI: 10.1152/ajpendo.00027.2025 Negative / Null Result ReportMedicine
Fernández-Ruiz M, Sánchez Moreno B, Santiago Almeda J et al. · 2023 · Transplant infectious disease : an official journal of the Transplantation Society
Previous studies have suggested that exposure to statins confers a protective effect in bloodstream infection (BSI) due to the anti-inflammatory and immunomodulatory properties attributed to these lipid-lowering drugs. Scarce evidence is…
View details →DOI: 10.1111/tid.14132 Negative / Null Result ReportOpen accessMedicine
Vilk Ayalon N, Omer Vilk N, Sompolinsky Y et al. · 2026 · Reproductive health
BACKGROUND: Frozen embryo transfer (FET) is widely used due to its safety and effectiveness. However, it has been associated with an increased risk of large for gestational age (LGA) infants. Proposed mechanisms include effects of embryo…
View details →DOI: 10.1186/s12978-026-02322-z Negative / Null Result ReportOpen accessMedicine
Agudelo LR, Yarmush G, Kumar S et al. · 2025 · Biology
Accelerating healing is a clinical goal in both acute and chronic non-healing skin wounds. We leveraged the public Recon database, which seeks to aggregate all of the metabolic pathways in the human body, to uncover whether increasing the…
View details →DOI: 10.3390/biology14060722 Negative / Null Result ReportMedicine
Sandoval-Alarcón S, Bäumle R, Castellucci HI · 2025 · Applied ergonomics
Dentists face various occupational demands that heighten their risk of developing work-related musculoskeletal disorders (WMSDs), including prolonged static postures, repetitive movements, unsuitable dental instruments, extended work…
View details →DOI: 10.1016/j.apergo.2025.104602 Negative / Null Result ReportOpen accessMedicine
Warren WG, Papagianni EP, Hale E et al. · 2022 · Frontiers in pharmacology
Endocannabinoid transmission is emerging as a target for treating anxiety-related disorders, given its regulation of fear extinction. Boosting anandamide levels via inhibition of its metabolism by fatty acid amide hydrolase (FAAH) can…
View details →DOI: 10.3389/fphar.2022.1082760 Negative / Null Result ReportOpen accessMedicine
Khalid E, Vijayan MM · 2025 · Frontiers in endocrinology
Introduction Although the Crh-Crhr1 system is the proximal trigger for the stressor-induced corticosteroid release, its role in initiating the fight-or-flight response to an acute stressor is unclear. We hypothesized that the Crh-Crhr1…
View details →DOI: 10.3389/fendo.2025.1650458 Negative / Null Result ReportOpen accessMedicine
Ouerghi N, Abassi W, Jebabli N et al. · 2026 · Sports (Basel, Switzerland)
The study investigated the effects of high-intensity interval training (HIIT) on cardiorespiratory fitness, hormonal, and psychological markers in adolescents. Twenty-eight healthy male adolescents were randomized to a HIIT group or a…
View details →DOI: 10.3390/sports14050209 Negative / Null Result ReportOpen accessMedicine
Toledo FV, Meletti JFA, Silva LMC et al. · 2026 · Brazilian journal of anesthesiology (Elsevier)
Introduction Surgery remains one of the most important treatments for breast cancer. In this context, the quality of postoperative recovery has become a key concern. Adequate control of acute pain is essential to optimize patient comfort…
View details →DOI: 10.1016/j.bjane.2026.844749 Negative / Null Result ReportMedicine
Salihu AT, Usman JS, Hill KD et al. · 2023 · Experimental brain research
The ability to control balance and prevent falls while carrying out daily life activities may require a predominantly controlled (cognitive) or automatic processing depending on the balance challenge, age, or other factors. Consequently,…
View details →DOI: 10.1007/s00221-023-06643-4 Negative / Null Result ReportOpen accessMedicine
Tatsuoka Y, Zembrzuski KJ, Natalini JG et al. · 2025 · Journal of clinical medicine
Background: Primary graft dysfunction (PGD) is the leading cause of early mortality after lung transplantation. Albumin is commonly used during lung transplantation to maintain intravascular volume while minimizing total intravenous fluid…
View details →DOI: 10.3390/jcm14217843 Negative / Null Result ReportOpen accessMedicine
Berge M, Hultgren O, Hugosson S et al. · 2026 · Scientific reports
There are large variations in how individual patients respond to allergen immunotherapy (AIT) against grass and/or birch allergy. There are currently no reliable biomarkers to predict which patients are likely to benefit from the…
View details →DOI: 10.1038/s41598-026-53894-6 Negative / Null Result ReportOpen accessMedicine
Mariani C, Al Dababsekh A, Carta F et al. · 2025 · Medicina (Kaunas, Lithuania)
Background and Objectives : Hemostasis during thyroid surgery is crucial to avoid postoperative complications, particularly hematoma, which can cause life-threatening airway compromise. Arista™ AH, a plant-based absorbable hemostatic…
View details →DOI: 10.3390/medicina61122209 Negative / Null Result ReportOpen accessMedicine
Jobson I, Vo NTN, Kujawinski E et al. · 2025 · Bioelectronic medicine
Background In glioblastoma (GBM) therapy research, tumour treating fields by the company Novocure™, have shown promise for increasing patient overall survival. When used with the chemotherapeutic agent temozolomide, they extend median…
View details →DOI: 10.1186/s42234-024-00164-3 