Derives geodesic ridge regularization and Riemannian Gibbs Process prior for feature-learning wide neural networks, generalizing kernel-regime results via function-space axiomatization.
hub Mixed citations
Deep residual learning for image recognition
Mixed citation behavior. Most common role is background (58%).
76
Pith papers citing it
Background
58% of classified citations
hub tools
citation-role summary
background 11
method 5
baseline 2
dataset 1
citation-polarity summary
representative citing papers
Develops a mean-field neural PDE model for transformer training, proves forward-pass well-posedness via function-space ODEs, derives conditional Wasserstein gradients, and shows global convergence of gradient flow under an NTK injectivity condition equivalent to linear independence of log-sum-exp fu
Introduces RAM-W600, the first public multi-task dataset of wrist conventional radiographs with instance segmentation annotations and Sharp/van der Heijde bone erosion scores for rheumatoid arthritis research.
The paper derives a posteriori error estimates for neural network depth adaptation by formulating training as an optimal control problem and using dual weighted residuals to insert layers where error is highest.
Introduces complement-aware submodular functions (CSI) that preserve structure between subset and complement for improved robust data selection.
MAPS provides 2618 validated 3D meshes and a controllable rendering pipeline to attribute vision model recognition failures to specific scene parameters, finding camera distance and elevation as the dominant failure factors across 20 tested models.
SPA unlocks patch-level features in CLIP for class-incremental learning via semantic-guided selection and optimal transport alignment with class descriptions, plus projectors and pseudo-feature replay to reduce forgetting.
Backdoors can be embedded in ResNet and ViT models as statistically indistinguishable latent directions, reducing cryptographic undetectability to an intractable hypothesis test over parameter distributions.
ImageAttributionBench is a benchmark dataset demonstrating that state-of-the-art image attribution methods lack robustness to image degradation and fail to generalize to semantically disjoint domains.
SubPopMark embeds verifiable subpopulation biases into distilled datasets via CVM and USTM optimization stages, allowing provenance inference through comparison of model output signatures against a reference behavior bank.
TCP-SSM conditions stable poles on visual tokens to explicitly control memory decay and oscillation in SSMs, cutting computation up to 44% while matching or exceeding accuracy on classification, segmentation, and detection.
GraphScan replaces geometric or coordinate-based scanning in Vision SSMs with learned local semantic graph routing, yielding SOTA results among such models on classification and segmentation tasks.
SSDA uses spectral magnitude alignment and structural-guided low-rank adaptation to close frequency and adjacency gaps when large vision models process time series rendered as images.
Intrinsic Muon provides closed-form linear maximization oracles on multiple Riemannian matrix manifolds for unitarily invariant norms, with convergence rates depending only on manifold dimension or rank.
Slingshot loss spikes are produced by low-precision arithmetic that breaks the zero-sum gradient constraint and drives exponential growth via Numerical Feature Inflation.
FGN is a positive semidefinite under-approximation of the multiclass GGN obtained by exact decomposition into true-vs-rest and within-competitor terms, exact for binary classification and implemented via matrix-free conjugate gradient on a whitened row-space system.
RAM-H1200 introduces a public dataset of 1,200 hand X-rays with whole-hand bone segmentation, pixel-level bone erosion masks, and joint-level SvdH scores for both erosion and narrowing to enable unified RA analysis.
In linear recurrent models, infinite-width signal propagation remains accurate only for depths t much smaller than sqrt(width n), with a critical regime at t ~ c sqrt(n) where finite-width effects emerge and dominate for larger t.
CNN classifiers work by holographic superposition and destructive interference in pixel space rather than selecting cleaned features, as proven by a new adjoint inversion framework that also yields a covariance-volume channel selection algorithm.
An acoustic dimer of two subwavelength scatterers can achieve unidirectional transverse scattering (transverse Kerker effect) while maintaining strong overall scattering.
FORGE benchmark shows domain-specific knowledge, not visual grounding, is the main bottleneck for MLLMs in manufacturing, with SFT on a 3B model delivering up to 90.8% relative accuracy improvement on held-out scenarios.
Unimodal model representations converge to a relational structure captured by the Indra representation via V-enriched Yoneda embedding, which is unique and structure-preserving and improves cross-model and cross-modal robustness when instantiated with angular distance.
ROI coding enables backdoor triggers to survive lossy compression by embedding malicious information into binary bitstreams via sample-specific or customized masks for both learned and traditional codecs.
Large loss spikes in SGD are polynomially likely and serve as the dominant mechanism for escaping sharp minima toward flatter solutions in the NTK regime.
citing papers explorer
-
Canonical Regularisation of Wide Feature-Learning Neural Networks
Derives geodesic ridge regularization and Riemannian Gibbs Process prior for feature-learning wide neural networks, generalizing kernel-regime results via function-space axiomatization.
-
Training Infinitely Deep and Wide Transformers
Develops a mean-field neural PDE model for transformer training, proves forward-pass well-posedness via function-space ODEs, derives conditional Wasserstein gradients, and shows global convergence of gradient flow under an NTK injectivity condition equivalent to linear independence of log-sum-exp fu
-
RAM-W600: A Multi-Task Wrist Dataset and Benchmark for Rheumatoid Arthritis
Introduces RAM-W600, the first public multi-task dataset of wrist conventional radiographs with instance segmentation annotations and Sharp/van der Heijde bone erosion scores for rheumatoid arthritis research.
-
An optimal control approach for neural network architecture adaptation with a posteriori error estimation
The paper derives a posteriori error estimates for neural network depth adaptation by formulating training as an optimal control problem and using dual weighted residuals to insert layers where error is highest.
-
Complement Submodular Information Measures for Balanced and Robust Data Selection
Introduces complement-aware submodular functions (CSI) that preserve structure between subset and complement for improved robust data selection.
-
MAPS: A Synthetic Dataset for Probing Vision Models in a Controlled 3D Scene Space
MAPS provides 2618 validated 3D meshes and a controllable rendering pipeline to attribute vision model recognition failures to specific scene parameters, finding camera distance and elevation as the dominant failure factors across 20 tested models.
-
Unlocking Patch-Level Features for CLIP-Based Class-Incremental Learning
SPA unlocks patch-level features in CLIP for class-incremental learning via semantic-guided selection and optimal transport alignment with class descriptions, plus projectors and pseudo-feature replay to reduce forgetting.
-
Backdoor Channels Hidden in Latent Space: Cryptographic Undetectability in Modern Neural Networks
Backdoors can be embedded in ResNet and ViT models as statistically indistinguishable latent directions, reducing cryptographic undetectability to an intractable hypothesis test over parameter distributions.
-
ImageAttributionBench: How Far Are We from Generalizable Attribution?
ImageAttributionBench is a benchmark dataset demonstrating that state-of-the-art image attribution methods lack robustness to image degradation and fail to generalize to semantically disjoint domains.
-
From Compression to Accountability: Harmless Copyright Protection for Dataset Distillation
SubPopMark embeds verifiable subpopulation biases into distilled datasets via CVM and USTM optimization stages, allowing provenance inference through comparison of model output signatures against a reference behavior bank.
-
TCP-SSM: Efficient Vision State Space Models with Token-Conditioned Poles
TCP-SSM conditions stable poles on visual tokens to explicitly control memory decay and oscillation in SSMs, cutting computation up to 44% while matching or exceeding accuracy on classification, segmentation, and detection.
-
Can Graphs Help Vision SSMs See Better?
GraphScan replaces geometric or coordinate-based scanning in Vision SSMs with learned local semantic graph routing, yielding SOTA results among such models on classification and segmentation tasks.
-
SSDA: Bridging Spectral and Structural Gaps via Dual Adaptation for Vision-Based Time Series Forecasting
SSDA uses spectral magnitude alignment and structural-guided low-rank adaptation to close frequency and adjacency gaps when large vision models process time series rendered as images.
-
Intrinsic Muon: Spectral Optimization on Riemannian Matrix Manifolds
Intrinsic Muon provides closed-form linear maximization oracles on multiple Riemannian matrix manifolds for unitarily invariant norms, with convergence rates depending only on manifold dimension or rank.
-
Grokking or Glitching? How Low-Precision Drives Slingshot Loss Spikes
Slingshot loss spikes are produced by low-precision arithmetic that breaks the zero-sum gradient constraint and drives exponential growth via Numerical Feature Inflation.
-
Fast Gauss-Newton for Multiclass Cross-Entropy
FGN is a positive semidefinite under-approximation of the multiclass GGN obtained by exact decomposition into true-vs-rest and within-competitor terms, exact for binary classification and implemented via matrix-free conjugate gradient on a whitened row-space system.
-
RAM-H1200: A Unified Evaluation and Dataset on Hand Radiographs for Rheumatoid Arthritis
RAM-H1200 introduces a public dataset of 1,200 hand X-rays with whole-hand bone segmentation, pixel-level bone erosion masks, and joint-level SvdH scores for both erosion and narrowing to enable unified RA analysis.
-
How Long Does Infinite Width Last? Signal Propagation in Long-Range Linear Recurrences
In linear recurrent models, infinite-width signal propagation remains accurate only for depths t much smaller than sqrt(width n), with a critical regime at t ~ c sqrt(n) where finite-width effects emerge and dominate for larger t.
-
Adjoint Inversion Reveals Holographic Superposition and Destructive Interference in CNN Classifiers
CNN classifiers work by holographic superposition and destructive interference in pixel space rather than selecting cleaned features, as proven by a new adjoint inversion framework that also yields a covariance-volume channel selection algorithm.
-
FingerEye: Learning Dexterous Manipulation with Continuous Vision-Tactile Sensing
An acoustic dimer of two subwavelength scatterers can achieve unidirectional transverse scattering (transverse Kerker effect) while maintaining strong overall scattering.
-
FORGE: Fine-grained Multimodal Evaluation for Manufacturing Scenarios
FORGE benchmark shows domain-specific knowledge, not visual grounding, is the main bottleneck for MLLMs in manufacturing, with SFT on a 3B model delivering up to 90.8% relative accuracy improvement on held-out scenarios.
-
The Indra Representation Hypothesis for Multimodal Alignment
Unimodal model representations converge to a relational structure captured by the Indra representation via V-enriched Yoneda embedding, which is unique and structure-preserving and improves cross-model and cross-modal robustness when instantiated with angular distance.
-
Inevitable Encounters: Backdoor Attacks Involving Lossy Compression
ROI coding enables backdoor triggers to survive lossy compression by embedding malicious information into binary bitstreams via sample-specific or customized masks for both learned and traditional codecs.
-
Large Spikes in Stochastic Gradient Descent: A Large-Deviations View
Large loss spikes in SGD are polynomially likely and serve as the dominant mechanism for escaping sharp minima toward flatter solutions in the NTK regime.
-
Spatially Localized Image Degradation Embeddings for Image Quality Assessment
SLIDE-IQA uses a dual-branch ViT with Threshold-Bounded Exclusion Mechanism for contrastive pretraining on localized degradations to boost sensitivity in NR-IQA while matching existing SSL benchmarks.
-
AnyMo: Geometry-Aware Setup-Agnostic Modeling of Human Motion in the Wild
AnyMo pre-trains a graph encoder on physics-simulated multi-placement IMU data and aligns full-body motion tokens with LLMs to enable zero-shot activity recognition, retrieval, and captioning across unseen datasets and setups.
-
A New Framework to Analyse the Distributional Robustness of Deep Neural Networks
A framework models DNN layer weight-activation interactions via Bernoulli distributions and uses class separation as a diagnostic proxy to quantify distributional robustness, tested on CIFAR-10 and ImageNet models.
-
MAM-CLIP: Vision-Language Pretraining on Mammography Atlases for BI-RADS Classification
Contrastive pretraining on mammography atlas image-text pairs improves BI-RADS classification F1 by 1-14% especially in low-label regimes, outperforming equivalent numbers of direct labels in some settings.
-
Perfect Parallelization in Mini-Batch SGD with Classical Momentum Acceleration
Classical momentum acceleration in mini-batch SGD for quadratics is proportional to batch size up to saturation, enabling perfect parallelization under minimal noise assumptions.
-
Concise and Logically Consistent Conformal Sets for Neuro-Symbolic Concept-Based Models
COCOCO is a conformal framework for NeSy-CBMs that jointly conformalizes concepts and labels, reconciles them via deduction-abduction revision, and satisfies consistency, coverage, and conciseness while retaining distribution-free guarantees.
-
SpecSem-Net: Integrating Spectral and Semantic Features for Robust AI-generated Video Detection
SpecSem-Net integrates Fourier-based spectral filtering with semantic-guided gated merging to detect AI-generated videos, reporting 87.25% accuracy on a new benchmark of five commercial generators and 95.59% on public datasets.
-
Wavelet Flow Matching for Multi-Scale Physics Emulation
Wavelet Flow Matching emulates multi-scale PDE-governed systems by transporting velocities directly in a hierarchical wavelet representation via U-Net, yielding improved long-horizon stability and spectral accuracy on fluid benchmarks.
-
STARS: Spike Tail-Aware Relational Synthesis for ANN-to-SNN Data-Free Knowledge Distillation
STARS augments BN-guided data synthesis with relational consistency alignment and tail-aware regularization for ANN-to-SNN data-free KD, reporting gains up to 4.6% on CIFAR-10 and 6.7% on CIFAR-100.
-
Temporal Sampling Frequency Matters: A Capacity-Aware Study of End-to-End Driving Trajectory Prediction
Smaller end-to-end autonomous driving models achieve optimal 3-second trajectory prediction accuracy at lower or intermediate temporal sampling frequencies, whereas larger VLA-style models perform best at the highest frequencies across Waymo, nuScenes, and PAVE datasets.
-
L2A: Learning to Accumulate Pose History for Accurate 3D Human Pose Estimation
A new method accumulates historical pose features across layers in a Transformer network to reach state-of-the-art 3D human pose estimation accuracy.
-
Penalty-Based First-Order Methods for Bilevel Optimization with Minimax and Constrained Lower-Level Problems
Penalty-based first-order methods find ε-KKT points in bilevel minimax problems with Õ(ε^{-4}) deterministic and Õ(ε^{-9}) stochastic oracle complexity, improving prior bounds for constrained lower-level cases via Lagrangian duality.
-
Robust stochastic first order methods in heavy-tailed noise via medoid mini-batch gradient sampling
R-SGD-Mini achieves O(1/T) convergence of expected squared gradient norm to a noise-dependent neighborhood in heavy-tailed settings by selecting the medoid gradient from M data chunks.
-
Inference-Time Attribute Distribution Alignment for Unconditional Diffusion
An optimal control formulation adds time-dependent perturbations to the reverse diffusion process to match target attribute distributions while preserving sample fidelity.
-
UniISP: A Unified ISP Framework for Both Human and Machine Vision
UniISP unifies ISP processing with a Hybrid Attention Module and Feature Adapter to produce images that are both visually pleasing for humans and informative for computer vision models.
-
When Labels Have Structure: Improving Image Classification with Hierarchy-Aware Cross-Entropy
Hierarchy-Aware Cross-Entropy improves image classification by incorporating class hierarchies into the loss through prediction aggregation and ancestral label smoothing, achieving mean accuracy gains of 4.66% in end-to-end training and 2.18% in linear probing.
-
Taming Outlier Tokens in Diffusion Transformers
Outlier tokens in DiTs are addressed with Dual-Stage Registers, which reduce artifacts and improve image generation on ImageNet and text-to-image tasks.
-
Height-Guided Projection Reparameterization for Camera-LiDAR Occupancy
HiPR improves 3D occupancy prediction by reparameterizing image-to-voxel projections using LiDAR-derived height priors to adapt sampling ranges to scene sparsity and height variations.
-
SignMuon: Communication-Efficient Distributed Muon Optimization
SignMuon merges majority-vote sign aggregation from signSGD with Muon's polar-factor steps to create a communication-efficient distributed optimizer that matches signSGD rates under symmetric noise and shows strong empirical results on CIFAR and nanoGPT.
-
Linear-Time Global Visual Modeling without Explicit Attention
Dynamic parameterization of standard layers can replace explicit attention for linear-time global visual modeling.
-
BOLT: Online Lightweight Adaptation for Preparation-Free Heterogeneous Cooperative Perception
BOLT is a 0.9M-parameter plug-and-play module that uses ego-as-teacher distillation on high-confidence predictions to align neighbor features online, raising AP@50 by up to 32.3 points over unadapted fusion while beating ego-only baselines on DAIR-V2X and OPV2V.
-
Fisher Decorator: Refining Flow Policy via a Local Transport Map
Fisher Decorator refines flow policies in offline RL via a local transport map and Fisher-matrix quadratic approximation of the KL constraint, yielding controllable error near the optimum and SOTA benchmark results.
-
SegMix:Shuffle-based Feedback Learning for Semantic Segmentation of Pathology Images
SegMix adaptively shuffles image patches and uses learning feedback to generate superior pseudo-semantic segmentation masks from image-level labels, outperforming prior methods on three pathology datasets.
-
Feed-Forward 3D Scene Modeling: A Problem-Driven Perspective
The paper proposes a problem-driven taxonomy for feed-forward 3D scene modeling that groups methods by five core challenges: feature enhancement, geometry awareness, model efficiency, augmentation strategies, and temporal-aware modeling.
-
StableTTA: Improving Vision Model Performance by Training-free Test-Time Adaptation Methods
StableTTA improves ImageNet-1K accuracy across 71 vision models by stabilizing logit aggregation under coherent-batch inference and enabling efficient single-forward-pass adaptation.
-
Solving and learning advective multiscale Darcian dynamics with the Neural Basis Method
The Neural Basis Method uses a predefined neural basis space and operator residual metric to deliver accurate single solves and fast parametric learning for multiscale Darcian dynamics.