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.

169,051 papers · 148 categories

Trend · papers per month

count % of corpus

7 results for “Celeb-A”

Compressing GANs using Knowledge Distillationcs.LG

New method compresses GANs by 1669:1 ratio on MNIST, 58:1 on CIFAR-10, 87:1 on Celeb-A.

problem Expensive GANs for low SWAP hardware and real-time applications.

method Knowledge distillation between a student and teacher GAN.

result Compressed GANs outperform standard training methods with a fixed parameter budget.

GAN compression knowledge distillation MNIST CIFAR-10 Celeb-A

2019-02-01abs ↗pdf ↗

Understanding Overparameterization in Generative Adversarial Networkscs.LG

This work analyzes how overparameterization aids GANs in reaching global saddle points.

problem Understanding the role of overparameterization in GANs for convergence to global saddle points.

method Theoretical and empirical analysis of overparameterized GANs with various architectures and datasets.

result GDA converges to a global saddle point in overparameterized GANs with certain assumptions.

Generative Adversarial Networks Overparameterization Global Convergence Non-convex Optimization CIFAR-10

2021-04-12abs ↗pdf ↗

Lifelong Generative Modelingstat.ML

Lifelong learning is the problem of learning multiple consecutive tasks in a sequential manner, where knowledge gained from previous tasks is retained and used to aid future learning over the lifetime of the learner. It is essential towards the development of intelligent machines that can adapt to their surroundings. I…

2017-05-27abs ↗pdf ↗

Sampling from Bayesian Neural Network Posteriors with Symmetric Minibatch Splitting Langevin Dynamicsstat.ML

New algorithm samples Bayesian neural networks for improved calibration.

problem Improving calibration of Bayesian neural networks.

method Symmetric Minibatch Splitting-UBU (SMS-UBU) algorithm.

result SMS-UBU provides better calibration performance than standard methods.

Bayesian neural networks sampling algorithms Langevin dynamics calibration stochastic weight averaging

2024-10-14abs ↗pdf ↗

Self-training Avoids Using Spurious Features Under Domain Shiftcs.LG

Self-training avoids spurious features in domain adaptation.

problem Domain shift with large differences between source and target domains.

method Entropy minimization on unlabeled target data, initialized with a source classifier.

result Entropy minimization avoids using spurious features in large domain shifts.

domain adaptation unsupervised learning entropy minimization spurious features label independence

2020-06-17abs ↗pdf ↗

Kanerva++: extending The Kanerva Machine with differentiable, locally block allocated latent memorycs.NE

New memory allocation scheme improves image generation performance.

problem Improving episodic and semantic memory representation in neural networks.

method Developed a hierarchical latent variable model with differentiable, locally block allocated latent memory.

result Improved conditional likelihood values on various datasets.

memory allocation latent variable model neural networks image generation Bayesian memory

2021-02-20abs ↗pdf ↗

Addressing GAN Training Instabilities via Tunable Classification Lossescs.LG

Improved GAN training stability through tunable classification losses.

problem Training instabilities in GANs.

method Reformulated GAN value function using class probability estimation (CPE) losses, defined

(α_D,α_G)

-GANs.

result Tuning

(α_D,α_G)

can alleviate training instabilities.

GANs classification losses training instabilities f-divergences α-loss

2023-10-27abs ↗pdf ↗