CIFAR-100
PulseAugur coverage of CIFAR-100 — every cluster mentioning CIFAR-100 across labs, papers, and developer communities, ranked by signal.
- instance of Class Incremental Learning 90%
- instance of CIFAR10-DVS: An Event-Stream Dataset for Object Classification 90%
- instance of Imagenet 1k 70%
- used by Imagenet 1k 70%
- used by ResNet-18 70%
- used by residual neural network 70%
- used by ResNet-50 70%
- instance of residual neural network 70%
- used by The Street View House Numbers Dataset 70%
- instance of ImageNet-100 70%
- used by Vision Transformers 70%
- instance of The Street View House Numbers Dataset 70%
20 day(s) with sentiment data
-
ZeroPur method offers training-free adversarial purification
Researchers have introduced ZeroPur, a novel method for adversarial purification that does not require additional training. This technique treats adversarial images as outliers from the natural image manifold and purifi…
-
New MAGIC-SSCIL framework improves semi-supervised incremental learning
Researchers have introduced MAGIC-SSCIL, a novel framework designed to address the significant challenge of Semi-supervised Class Incremental Learning (SSCIL) in neural networks, particularly in scenarios where past dat…
-
FEAST framework enhances federated learning for diverse client resources
Researchers have introduced FEAST, a novel framework for federated learning designed to accommodate clients with varying computational resources. FEAST trains a single elastic model, or "supernet," that can be adapted t…
-
New label granularity skew challenge identified in federated learning
Researchers have introduced a new challenge in federated learning called label granularity skew, where clients in a hierarchical image classification task provide labels at varying levels of detail. To address this, the…
-
Biologically-inspired D-SNN architecture enhances efficiency and transparency
Researchers have developed a Decomposable Spiking Neural Network (D-SNN) that mimics biological neural systems by isolating classification pathways into independent experts, thus avoiding global entanglement. This modul…
-
Study compares feature-based vs. logit-based knowledge distillation
A new study on arXiv investigates knowledge distillation techniques, specifically comparing feature-based methods against logit-based distillation across different student model architectures. Researchers found that whi…
-
New model quantitatively selects optimal transfer learning datasets
Researchers have developed TLDChoiceNet, a novel model designed to quantitatively select the optimal transfer learning dataset for image classification tasks. This system aims to address the lack of a systematic method …
-
New research explores advanced federated learning techniques · 10 sources tracked
Multiple research papers published on arXiv in August 2026 introduce novel approaches to enhance federated learning (FL) and decentralized FL. These methods address challenges such as modality missingness in multimodal …
-
New Transformer Architecture Optimizes Self-Supervised Learning
Researchers have developed an attention-only white-box Transformer model by integrating the LeJEPA self-supervised learning framework with optimization algorithms. This approach optimizes the sparse rate reduction objec…
-
SpecDrop introduces parameter-free routing for specialized AI models
Researchers have introduced SpecDrop, a novel parameter-free routing method for Mixture of Experts (MoE) models that leverages category labels for specialization. Unlike traditional MoE approaches that rely on learned r…
-
New methods improve neural network quantization efficiency and accuracy
Researchers have developed new methods for neural network quantization, a process that reduces the memory and computational requirements of AI models. The first paper introduces BaKron, an efficient solver that uses Kro…
-
New KD method enhances action recognition model compression
Researchers have developed a novel Channel-wise Dynamic Knowledge Distillation (KD) approach called ASCD KD to improve the compression of large action recognition models. This method addresses limitations in existing KD…
-
New analysis quantifies SAM's bias toward flat minima
Researchers have analyzed the implicit bias of Sharpness-Aware Minimization (SAM) in improving model generalization. Their linear stability analysis reveals a quantitative relationship between SAM's perturbation radius …
-
New AOS-R system optimizes deep learning training by switching optimizers
Researchers have developed AOS-R, a novel adaptive optimizer switching system designed to improve deep network training efficiency and generalization. This system monitors six online gradient-space signals to dynamicall…
-
LLMs drive neural architecture search with new methods for code and mobile deployment
Two new research papers explore the use of Large Language Models (LLMs) in Neural Architecture Search (NAS). The first paper, 'GraphIR', introduces an intermediate representation to bridge the gap between executable neu…
-
EulerLoRA enhances parameter-efficient fine-tuning with stochasticity
Researchers have developed EulerLoRA, a novel extension of the Low-Rank Adaptation (LoRA) technique for parameter-efficient fine-tuning. Unlike standard LoRA, EulerLoRA introduces stochasticity to generate multiple pred…
-
AI models fail calibration on unseen subtypes, research finds
A new research paper explores the concept of subtype robustness in AI models, focusing on whether models remain calibrated (i.e., their confidence matches their accuracy) when presented with data from fine-grained subty…
-
New method TOOD improves out-of-distribution detection in continual learning
A new paper introduces TOOD, a method designed to improve out-of-distribution (OOD) detection in continual learning systems. The research identifies two key issues: the "Confidence Gap" where energy-based detectors see …
-
New KAN Compression and Binary KAN Restoration Techniques Unveiled
Researchers have developed SparseKAN, a method to compress Kolmogorov-Arnold Networks (KANs) by reducing basis functions, neurons, and numerical precision. This approach aims to make KANs more efficient by removing redu…
-
New dual-teacher method boosts DNN robustness against adversarial attacks
Researchers have developed a new method to improve the robustness and accuracy of deep neural networks against adversarial attacks. This approach extends the Information Bottleneck Distillation (IBD) framework by incorp…