Research
On-device research index

arXiv research

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.

168,786 papers · 148 categories

Trend · papers per month

12.5%25.0%37.5%50.0% · Sep 199319922001200920172026
48 results for parallel diffusion

New method solves blind inverse problems by optimizing both operator and image parameters.

problem Solving blind inverse problems with known forward operator.
method Parallel reverse diffusion guided by gradients from intermediate stages.
result State-of-the-art performance on blind deblurring and imaging through turbulence.

The paper develops stochastic methods on geometric spaces for transformations.

problem Existence and uniqueness of stochastic processes on geometric spaces.
method Stochastic parallel transport and equivariant diffusions on the group of diffeomorphisms.
result Existence and uniqueness of stochastic parallel transport and equivariant diffusions.

Two parallel samplers enhance image quality in limited denoising steps.

problem Limited denoising steps in diffusion models reduce image quality.
method Two parallel samplers denoise at successive times, integrating their information.
result Two parallel samplers improve image quality compared to a single sampler.

ARDMs are a new model class for autoregressive diffusion that generalize existing models and can compress data efficiently.

problem Efficient data compression and generation.
method Autoregressive Diffusion Models (ARDMs) that generalize existing autoregressive models and discrete diffusion models.
result ARDMs require significantly fewer steps for compression compared to discrete diffusion models.

New sampling method improves efficiency for diffusion models.

problem Efficient sampling from arbitrary smooth distributions in polynomial time.
method Randomized midpoint method for log-concave sampling.
result Achieves best known dimension dependence (O~(d5/12)\widetilde O(d^{5/12})) for total variation distance.

New method speeds up diffusion models inference to sub-linear time.

problem Efficient inference of diffusion models for high-dimensional data.
method Parallel sampling with Picard iterations within blocks.
result Achieves sub-linear time complexity of O~(polylogd)\widetilde{\mathcal{O}}(\mathrm{poly} \log d).

Parallel score matching accelerates DPM training and improves density estimation.

problem Extended training periods and limited modeling flexibility in DPMs.
method Partitioning the learning task into independent time sub-intervals and modeling the score at each time point separately.
result Significant acceleration of training process and improved density estimation performance.

PTSD improves neural samplers by combining diffusion models and PT, enhancing efficiency.

problem Efficiency and correlation issues in neural samplers compared to PT.
method Sequential training of diffusion models across temperatures, combining high-temperature models for approximate lower-temperature samples.
result Significantly improved target evaluation efficiency, outperforming diffusion-based samplers.

DOS improves language model generation by considering inter-token dependencies.

problem Lack of sequence-level information and inter-token dependencies in existing decoding strategies.
method Dependency-Oriented Sampler (DOS) that uses attention matrices to approximate inter-token dependencies.
result DOS consistently achieves superior performance on code generation and mathematical reasoning tasks.

DreamPropeller accelerates text-to-3D generation by 4.7x with minimal loss in quality.

problem Long generation times in text-to-3D generation algorithms degrade user experience.
method DreamPropeller uses Picard iterations generalized for non-ODE paths to accelerate parallel sampling.
result Empirically achieves up to 4.7x speedup with negligible quality loss.

Researchers interpret SGD using diffusion metrics for clearer geometric understanding.

problem Elusiveness of geometrical significance in stochastic gradient descent.
method Study a deterministic model with geodesics of diffusion metrics.
result Establishes parallel with General Relativity models.

Diffusion models generate music sequences without autoregressive loops.

problem Generating music sequences from symbolic data using diffusion models.
method Parameterize discrete symbolic data in continuous latent space, train diffusion model, generate sequences through reverse process.
result Strong unconditional generation and post-hoc conditional infilling compared to autoregressive models.

New method adapts DLMs to intrinsic data dependence without prior knowledge.

problem Understanding how unmasking schedules affect DLM generation quality.
method Adapts unmasking schedule to target data distribution's dependence structure.
result Sampling convergence guarantees improve for low-complexity distributions.

Localized diffusion models reduce training complexity by exploiting low-dimensional structure.

problem Training diffusion models is computationally expensive due to the curse of dimensionality.
method Localized neural networks and localized score matching loss to estimate low-dimensional score functions.
result Localized diffusion models can circumvent the curse of dimensionality with reduced sample complexity.

New diffusion models can generate text faster than traditional autoregressive models.

problem The slow generation time of autoregressive models.
method Theoretical analysis of diffusion language models using information theory.
result The sampling error in diffusion models decays with fewer iterations than the text sequence length.

Theoretical analysis shows MDMs can be efficient but not for all metrics.

problem Understanding the efficiency-accuracy trade-off of diffusion language models.
method Theoretical analysis of Masked Diffusion Model (MDM) using perplexity and sequence error rate as metrics.
result MDM achieves near-optimal perplexity but requires linear scaling for sequence error rate, highlighting efficiency-accuracy trade-offs.

Latent Block-Diffusion Temporal Point Processes (LBDTPP) is a semi-autoregressive framework for generating asynchronous event sequences.

problem Generating asynchronous event sequences
method Latent Block-Diffusion Temporal Point Processes
result Outperforms state-of-the-art TPP baselines in both unconditional and conditional generation tasks

Graph-based approach predicts stock trends using dynamic multi-relational graphs.

problem Predicting future stock movements in complex, time-evolving stock relationships.
method Dynamic multi-relational stock graphs, stochastic diffusion process, parallel retention.
result Outperforms state-of-the-art baselines in stock trend forecasting.

C-DPS improves diffusion posterior sampling for inverse problems without projection or likelihood approximation.

problem Inaccurate and unstable solutions in inverse problems due to complex or high-noise conditions.
method C-DPS introduces a forward stochastic process in measurement space evolving in parallel with data-space diffusion, leading to a closed-form posterior.
result C-DPS consistently outperforms existing methods across multiple inverse problem benchmarks.

This research explores discrete diffusion models for natural language generation.

problem Challenges in applying diffusion models to discrete data, especially natural language.
method Investigates Discrete Denoising Diffusion Probabilistic Model (D3PM) and compares it with autoregressive models.
result Discrete diffusion models achieve better processing speed than autoregressive models.

In this paper we obtain a splitting theorem for the symmetric diffusion operator Δφ=Δ<φ,>Δ_φ=Δ-\left<\nablaφ,\nabla \right> and a non-constant C3C^3 function ff in a complete Riemannian manifold MM, under the assumptions that the Ricci curvature associated with ΔφΔ_φ satisfies Ricφ(f,f)0{\rm Ric}_φ(\nabla f,\nabla f)\ge 0, that $|…

2015-02-01abs ↗pdf ↗

Generative diffusion models improve financial LOB simulation and forecasting.

problem High noise and complexity in financial LOB data makes deep generative models ineffective.
method Convert LOB data to images, apply diffusion models with inpainting for long-term sequence generation.
result Our method achieves state-of-the-art performance on LOB-Bench, improving coherence over local details.

A hybrid model combines diffusion and neural operator methods for stress prediction in hyperelastic materials.

problem Challenges in predicting stress fields in hyperelastic materials with complex microstructures.
method A hybrid surrogate framework combining a conditional denoising diffusion probabilistic model (cDDPM) and a modified DeepONet.
result The hybrid model consistently outperforms traditional methods by one to two orders of magnitude.

The numerical solution of large-scale PDEs, such as those occurring in data-driven applications, unavoidably require powerful parallel computers and tailored parallel algorithms to make the best possible use of them. In fact, considerations about the parallelization and scalability of realistic problems are often criti…

2017-05-10abs ↗pdf ↗

New RL algorithm GDPO improves DLM reasoning efficiency.

problem Adapting RL to DLMs for efficient, unbiased likelihood estimation.
method Group Diffusion Policy Optimization (GDPO) using semi-deterministic Monte Carlo.
result GDPO outperforms existing methods on math, reasoning, and coding benchmarks.

New method uses zeroth-order queries to approximate proximal sampling efficiently.

problem Approximating proximal sampling with zeroth-order information.
method Direct simulation of heat flow dynamics, treating intermediate distribution as Gaussian mixture.
result Inherits exponential convergence under isoperimetric conditions, avoids rejection sampling.

Warped product metrics are a class of Riemannian metrics on cross products B×FB \times F which have been well studied and provide a rich set of examples. In this paper we consider shrinking gradient Ricci solitons which are warped product metrics. We prove that if the curvature of the metric is bounded and the base BB

2016-02-02abs ↗pdf ↗

Method locates equilibria on unknown Riemannian manifolds using iterative sampling and parallel transport.

problem Locating equilibria on unknown Riemannian manifolds defined by point-clouds.
method Iterative sampling, parallel transport, and generalized isoclines.
result Algorithm reliably locates equilibria of dynamical systems on unknown manifolds.

The aim of this paper is to develop a general framework for training neural networks (NNs) in a distributed environment, where training data is partitioned over a set of agents that communicate with each other through a sparse, possibly time-varying, connectivity pattern. In such distributed scenario, the training prob…

2016-10-24abs ↗pdf ↗

This paper applies reactor theory to supply chain management.

problem Maintaining optimal item delivery and collection ratios in supply chains.
method Translating neutron transport and diffusion theory to supply chain management, introducing analogy factors and interactors.
result A deterministic model for supply chain optimization.