CAML meta-learns a progressively refined inductive bias from active-learning queries to improve robustness to spurious correlations, reporting accuracy gains on minority groups across several benchmarks.
Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization,
29 Pith papers cite this work, alongside 5,572 external citations. Polarity classification is still indexing.
29
Pith papers citing it
5,572
external citations · OpenAlex
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
eX2L improves robustness to distribution shifts by penalizing similarity between Grad-CAM maps of a label classifier and a confounder classifier, reaching new SOTA average and worst-group accuracy on the Spawrious benchmark.
A training-free method using Fourier-parameterized star-convex contours optimized via gradients to generate compact, faithful visual attributions for image classifiers on benchmarks like ImageNet.
Phoneme-aligned Grad-CAM on a WavLM-CNN detector reveals significant attack- and speaker-dependent importance of vowels, fricatives and pauses for spoof vs bona-fide decisions on ASVspoof 5.
MICViT outperforms CNN and transformer baselines on brain age prediction from multimodal 3D MRI by combining modality-specific and cross-modal local/global attention across three heterogeneous datasets.
CascadeFormer tapers Transformer width with depth based on gradient fan-in asymmetry to match uniform baselines in perplexity while cutting latency 8.6%.
SimPhysNet achieves 96.06% accuracy classifying laser welding penetration states using self-supervised contrastive learning with a physics-informed neural network and prototypical networks on only 200 labeled images.
Wearable accelerometry, EDA, and temperature data from 9 students with profound autism, processed with fine-tuned foundation models, enables prediction of challenging behavior episodes up to 10 minutes in advance at AUC-ROC 0.78 in actual classroom sessions.
Transfer learning on a new clinical gait dataset shows selective freezing of low-level features in pretrained models yields stable frailty classification, with model attention aligning to lower-limb biomechanics.
CAAP produces patch attributions in ViTs by direct activation patching on intermediate layers to measure causal contribution to the target class score.
WaveDetect reformulates machine-generated text detection as a time-frequency signal processing task by applying continuous wavelet transform to token probability sequences to reveal spectral fingerprints.
DAR replaces GAP with an attention-based aggregation module retrained jointly with the classifier head to disentangle core from spurious features and outperforms DFR on multiple datasets.
Modified feedback alignment in convolutional networks produces representations geometrically aligned with backpropagation on CIFAR-10.
H-SemiS decomposes multi-class KOA severity grading into binary sub-tasks in a semi-supervised setup with self-supervision and quantum-inspired mixing, outperforming baselines on two multi-class and two binary datasets.
A novel algorithm learns sets of optimal quantile regression trees to predict full conditional distributions interpretably and efficiently.
Lightweight multi-task models using Gram matrices and PatchGAN-style architectures detect 53 weather classes from RGB images with F1 scores above 96% internally and 78% zero-shot externally, supported by a new 503k-image dataset.
Models predicting human authenticity judgments produce inconsistent attribution maps across architectures, showing that explanations are non-identifiable.
CLEAR-HPV restructures the latent space of attention-based MIL models to discover 10 label-free morphologic concepts that preserve slide-level HPV prediction performance and generalize across TCGA-HNSCC, TCGA-CESC, and CPTAC-HNSCC datasets.
BHB treatment restores autophagy, mitochondrial turnover, and vesicle morphology in C99-expressing Drosophila neurons via a VPS35-dependent mechanism.
Hybrid RL-PID controllers track angle of attack better and show greater robustness than PID alone within a defined operational envelope for re-entry attitude control.
Deep learning models on standardized 2D CT projections of pelvis and skull from 141 cadavers reach 95.65% patient-level accuracy for biological sex determination.
Dual-stream EEG decoder separates identity and orientation to support 3D reconstruction from neural signals via circular regression and conditioned diffusion.
A fine-tuned ViT on 8493 SEM images classifies fracture causes in zirconia-toughened alumina at 0.907 accuracy and 0.888 macro-F1, with comparable performance at 50x versus higher magnifications.
IncepDeHazeGAN is a GAN with Inception blocks and multi-layer feature fusion that claims state-of-the-art single-image dehazing performance on satellite datasets.
citing papers explorer
-
Cumulative Meta-Learning from Active Learning Queries for Robustness to Spurious Correlations
CAML meta-learns a progressively refined inductive bias from active-learning queries to improve robustness to spurious correlations, reporting accuracy gains on minority groups across several benchmarks.
-
eXplaining to Learn (eX2L): Regularization Using Contrastive Visual Explanation Pairs for Distribution Shifts
eX2L improves robustness to distribution shifts by penalizing similarity between Grad-CAM maps of a label classifier and a confounder classifier, reaching new SOTA average and worst-group accuracy on the Spawrious benchmark.
-
Extremal Contours: Gradient-driven contours for compact visual attribution
A training-free method using Fourier-parameterized star-convex contours optimized via gradients to generate compact, faithful visual attributions for image classifiers on benchmarks like ImageNet.
-
Why Do You Say It Like That? A Phoneme-Level Framework for Explainable Speech Deepfake Detection
Phoneme-aligned Grad-CAM on a WavLM-CNN detector reveals significant attack- and speaker-dependent importance of vowels, fricatives and pauses for spoof vs bona-fide decisions on ASVspoof 5.
-
Modeling Local, Global, and Cross-Modal Context in Multimodal 3D MRI
MICViT outperforms CNN and transformer baselines on brain age prediction from multimodal 3D MRI by combining modality-specific and cross-modal local/global attention across three heterogeneous datasets.
-
CascadeFormer: Depth-Tapered Transformers Motivated by Gradient Fan-in Asymmetry
CascadeFormer tapers Transformer width with depth based on gradient fan-in asymmetry to match uniform baselines in perplexity while cutting latency 8.6%.
-
A welding penetration prediction model for laser welding process based on self-supervised learning using physics-informed neural networks
SimPhysNet achieves 96.06% accuracy classifying laser welding penetration states using self-supervised contrastive learning with a physics-informed neural network and prototypical networks on only 200 labeled images.
-
Prediction of Challenging Behaviors Associated with Profound Autism in a Classroom Setting Using Wearable Sensors
Wearable accelerometry, EDA, and temperature data from 9 students with profound autism, processed with fine-tuned foundation models, enables prediction of challenging behavior episodes up to 10 minutes in advance at AUC-ROC 0.78 in actual classroom sessions.
-
The Gait Signature of Frailty: Transfer Learning based Deep Gait Models for Scalable Frailty Assessment
Transfer learning on a new clinical gait dataset shows selective freezing of low-level features in pretrained models yields stable frailty classification, with model attention aligning to lower-limb biomechanics.
-
Causal Attribution via Activation Patching
CAAP produces patch attributions in ViTs by direct activation patching on intermediate layers to measure causal contribution to the target class score.
-
WaveDetect: Robust Framework for Machine-Generated Text Detection via Wavelet Transform
WaveDetect reformulates machine-generated text detection as a time-frequency signal processing task by applying continuous wavelet transform to token probability sequences to reveal spectral fingerprints.
-
Deep Attention Reweighting: Post-Hoc Attention-Based Feature Aggregation in CNNs for Disentangling Core and Spurious Features under Spurious Correlations
DAR replaces GAP with an attention-based aggregation module retrained jointly with the classifier head to disentangle core from spurious features and outperforms DFR on multiple datasets.
-
Biological Plausibility and Representational Alignment of Feedback Alignment in Convolutional Networks
Modified feedback alignment in convolutional networks produces representations geometrically aligned with backpropagation on CIFAR-10.
-
H-SemiS: Hierarchical Fusion of Semi and Self-Supervised Learning for Knee Osteoarthritis Severity Grading
H-SemiS decomposes multi-class KOA severity grading into binary sub-tasks in a semi-supervised setup with self-supervision and quantum-inspired mixing, outperforming baselines on two multi-class and two binary datasets.
-
Interpretable Quantile Regression by Optimal Decision Trees
A novel algorithm learns sets of optimal quantile regression trees to predict full conditional distributions interpretably and efficiently.
-
Heuristic Style Transfer for Real-Time, Efficient Weather Attribute Detection
Lightweight multi-task models using Gram matrices and PatchGAN-style architectures detect 53 weather classes from RGB images with F1 scores above 96% internally and 78% zero-shot externally, supported by a new 503k-image dataset.
-
Non-identifiability of Explanations from Model Behavior in Deep Networks of Image Authenticity Judgments
Models predicting human authenticity judgments produce inconsistent attribution maps across architectures, showing that explanations are non-identifiable.
-
CLEAR-HPV: Interpretable concept discovery for human-papillomavirus-associated morphology in whole-slide histology
CLEAR-HPV restructures the latent space of attention-based MIL models to discover 10 label-free morphologic concepts that preserve slide-level HPV prediction performance and generalize across TCGA-HNSCC, TCGA-CESC, and CPTAC-HNSCC datasets.
-
beta Hydroxybutyrate remodels the C99 interactome and coincides with restored organelle homeostasis in a Drosophila Alzheimers model
BHB treatment restores autophagy, mitochondrial turnover, and vesicle morphology in C99-expressing Drosophila neurons via a VPS35-dependent mechanism.
-
Deep Reinforcement Learning for Spacecraft Attitude Control During Atmospheric Re-Entry
Hybrid RL-PID controllers track angle of attack better and show greater robustness than PID alone within a defined operational envelope for re-entry attitude control.
-
Biological Sex Determination in Cadavers Using Deep Learning Algorithms from Computed Tomography Images of Pelvis and Skull
Deep learning models on standardized 2D CT projections of pelvis and skull from 141 cadavers reach 95.65% patient-level accuracy for biological sex determination.
-
Dual-Stream EEG Decoding for 3D Visual Perception
Dual-stream EEG decoder separates identity and orientation to support 3D reconstruction from neural signals via circular regression and conditioned diffusion.
-
Low-Magnification SEM May Suffice: Interpretable Deep Learning for Multi-Scale Fracture-Cause Classification in Zirconia-Toughened Alumina
A fine-tuned ViT on 8493 SEM images classifies fracture causes in zirconia-toughened alumina at 0.907 accuracy and 0.888 macro-F1, with comparable performance at 50x versus higher magnifications.
-
IncepDeHazeGAN: Novel Satellite Image Dehazing
IncepDeHazeGAN is a GAN with Inception blocks and multi-layer feature fusion that claims state-of-the-art single-image dehazing performance on satellite datasets.
-
Toward Aristotelian Medical Representations: Backpropagation-Free Layer-wise Analysis for Interpretable Generalized Metric Learning on MedMNIST
A-ROM delivers competitive MedMNIST performance via pretrained ViT metric spaces, a concept dictionary, and kNN without backpropagation or fine-tuning, framed as interpretable few-shot learning under the Platonic Representation Hypothesis.
-
Cross-Architectural Mixture-of-Experts with Adaptive Soft Routing for Plant Leaf Disease Classification
A soft-gated MoE combining EfficientNet-B0, DenseNet-121, and Swin-Tiny reports 92.62% F1-score on an imbalanced potato leaf disease dataset, outperforming single models by 5%.
-
Comparing Post-Hoc Explainable AI Methods for Interpreting Black-Box EEG Models in Depression Detection
Gradient- and perturbation-based XAI methods show substantial agreement on frontal, temporal, and posterior EEG regions for an InceptionTime MDD classifier, while DeepSHAP differs, with overall partial convergence and method-dependent variability.
-
Evaluating Local Explainability Metrics for Machine Learning Models on Tabular Data
Benchmark of local explainability methods on tabular data finds explanation quality driven primarily by dataset complexity rather than model predictive performance.
-
Interpreting "Interpretability" and Explaining "Explainability" in Machine Learning in Physics
The paper defines interpretability as model structural transparency and explainability as scientific content mapping, discusses their trade-offs, and frames both as deliberate modeling choices for ML in physics.