A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
New method uses LLMs to generate detailed scientific hypotheses.
Convolutional GANs favor low spatial frequencies, affecting fine detail generation.
Residual Prior Diffusion integrates coarse latent priors with diffusion models for better generative tasks.
Many practical techniques for probabilistic inference require a sequence of distributions that interpolate between a tractable distribution and an intractable distribution of interest. Usually, the sequences used are simple, e.g., based on geometric averages between distributions. When models are expressed as probabili…
Understanding the power of depth in feed-forward neural networks is an ongoing challenge in the field of deep learning theory. While current works account for the importance of depth for the expressive power of neural-networks, it remains an open question whether these benefits are exploited during a gradient-based opt…
This paper circulated previously in a draft version. Now, upon general request, it is about time to distribute the more detailed (and much longer) version. The main technical issues revolve around the fine structure of the compactification of the moduli spaces of flow lines and the obstruction bundle technique, with re…
In this study, we propose the integration of competitive learning into convolutional neural networks (CNNs) to improve the representation learning and efficiency of fine-tuning. Conventional CNNs use back propagation learning, and it enables powerful representation learning by a discrimination task. However, it require…
We describe a new training methodology for generative adversarial networks. The key idea is to grow both the generator and discriminator progressively: starting from a low resolution, we add new layers that model increasingly fine details as training progresses. This both speeds the training up and greatly stabilizes i…
AI model automates financial investment research tasks.
We provide a detailed characterization of the optimal consumption stream for the additive habit-forming utility maximization problem, in a framework of general discrete-time incomplete markets and random endowments. This characterization allows us to derive the monotonicity and concavity of the optimal consumption as a…
DIF extends NF with stochastic discrete latent variables for better density estimation.
EQD model improves domain-specific QA by 0.6% to 10.5%.
MuSiCNet tackles irregularly sampled multivariate time series by treating them as a hierarchy of relatively regular series.
We provide an online RLHF workflow for large language models.
Recent sparse MRI reconstruction models have used Deep Neural Networks (DNNs) to reconstruct relatively high-quality images from highly undersampled k-space data, enabling much faster MRI scanning. However, these techniques sometimes struggle to reconstruct sharp images that preserve fine detail while maintaining a nat…
In many contexts, we have access to aggregate data, but individual level data is unavailable. For example, medical studies sometimes report only aggregate statistics about disease prevalence because of privacy concerns. Even so, many a time it is desirable, and in fact could be necessary to infer individual level chara…
Powerful generative adversarial networks (GAN) have been developed to automatically synthesize realistic images from text. However, most existing tasks are limited to generating simple images such as flowers from captions. In this work, we extend this problem to the less addressed domain of face generation from fine-gr…
Various semigroups of noninvertible supermatrices of the special (antitriangle) shape having nilpotent Berezinian which appear in supersymmetric theories are defined and investigated. A subset of them continuously represents left and right zero semigroups and rectangular bands. The ideal properties of higher order rect…
Fast and accurate MRI image reconstruction from undersampled data is crucial in clinical practice. Deep learning based reconstruction methods have shown promising advances in recent years. However, recovering fine details from undersampled data is still challenging. In this paper, we introduce a novel deep learning bas…
Deep neural networks paved the way for significant improvements in image visual categorization during the last years. However, even though the tasks are highly varying, differing in complexity and difficulty, existing solutions mostly build on the same architectural decisions. This also applies to the selection of acti…
Paper introduces a flow-based framework for representation learning.
DiffDenoise preserves fine structures in medical images using conditional diffusion models.
This paper simplifies fine-tuning for small LLMs, reducing barriers for developers.
ECC Analyzer uses LLMs to predict stock volatility from ECCs.
To improve the compressive sensing MRI (CS-MRI) approaches in terms of fine structure loss under high acceleration factors, we have proposed an iterative feature refinement model (IFR-CS), equipped with fixed transforms, to restore the meaningful structures and details. Nevertheless, the proposed IFR-CS still has some …
This paper proposes a subspace decomposition method based on an over-complete dictionary in sparse representation, called "Sparse Signal Subspace Decomposition" (or 3SD) method. This method makes use of a novel criterion based on the occurrence frequency of atoms of the dictionary over the data set. This criterion, wel…
We study the geometry of the Thurston metric on the Teichmüller space of hyperbolic structures on a surface . Some of our results on the coarse geometry of this metric apply to arbitrary surfaces of finite type; however, we focus particular attention on the case where the surface is a once-punct…
The paper introduces a Hessian-based method to improve generalization in fine-tuned deep neural networks.
Recently, there has been great interest in connections between continuous-time dynamical systems and optimization methods, notably in the context of accelerated methods for smooth and unconstrained problems. In this paper we extend this perspective to nonsmooth and constrained problems by obtaining differential inclusi…
Unsupervised pre-training improves model generalization, but lacks theoretical understanding.
The paper develops a theory linking pretraining and fine-tuning in neural networks.
Adjoint Matching improves flow and diffusion models with reward fine-tuning.
New math for deep learning tackles key questions about neural networks.
Motivated by recently published methods using frequency decompositions of convolutions (e.g. Octave Convolutions), we propose a novel convolution scheme to stabilize the training and reduce the likelihood of a mode collapse. The basic idea of our approach is to split convolutional filters into additive high and low fre…
Self-play fine-tuning improves diffusion models for text-to-image generation.
Improved code translation by preserving structure with composed fine-tuning.
LP-FT improves personalized model training in FL by balancing generalization and personalization.
New method quantifies uncertainty in fine-tuned LLMs using LoRA ensembles.
Characterizes geometric actions on graphs with flexible stabilizers.
Transformers fine-tuned on synthetic data boost tabular data classification performance.
Abstract: Proves generic torus diffeomorphisms act parabolically and non-properly on fine curve graph and have generalized rotation sets.
Human matting, high quality extraction of humans from natural images, is crucial for a wide variety of applications. Since the matting problem is severely under-constrained, most previous methods require user interactions to take user designated trimaps or scribbles as constraints. This user-in-the-loop nature makes th…
BERT fine-tuning is unstable due to optimization issues, not forgetting or dataset size.
Simplified trust region method reduces representation change during fine-tuning.
Automorphisms of fine graphs for surfaces and tori are studied.
Fine-tuning with pre-training data improves performance.
This research improves calorimeter simulations by creating a faster model.
Fine-grained pretraining improves neural network's ability to learn rare features.