MeshTok uses AMR-inspired adaptive multiscale tokenization to improve the efficiency-accuracy trade-off of Transformer models for PDEs over uniform-grid baselines.
Title resolution pending
13 Pith papers cite this work. Polarity classification is still indexing.
13
Pith papers citing it
citation-role summary
background 3
method 1
citation-polarity summary
representative citing papers
LG-CoTrain, an LLM-guided co-training method, outperforms classical semi-supervised baselines for crisis tweet classification in low-resource settings with 5-25 labeled examples per class.
EBMs trained with non-persistent short runs reproduce empirical data statistics via a precise dynamical process, not the equilibrium measure.
Develops a local tangent-space rate-distortion theory and eigenspace-overlap diagnostic showing when physics-aligned compression necessarily degrades standard fidelity due to misaligned sensitivity directions.
A single energy-based model trained on LAPD plasma data enables diagnostic reconstruction, inverse inference of probe position, conditional trend sampling, and unconditional mode reproduction for potential anomaly detection.
PIT uses a neural autoencoder with a differentiable physics module and a new Physics-Informed Landmark Loss to track single particles in video, achieving sub-pixel accuracy in supervised and unsupervised modes.
The authors introduce dRVFL and edRVFL frameworks that stack RVFL layers with fixed random weights and closed-form outputs, reporting superior benchmark performance when combined with sparse-pretrained RVFL.
A multi-block attention neural network reduces pilot overhead by 87% and NMSE by 51% at 10 dB SNR for cascaded channel estimation in IRS-assisted mmWave MIMO-OFDM systems.
PixelFlowCast delivers high-fidelity precipitation nowcasts from radar sequences using a latent-free Pixel Mean Flows predictor guided by a deterministic coarse stage and KANCondNet features.
Timestep embeddings are redundant in diffusion models under certain conditions, with time-agnostic variants matching or exceeding conditioned models on FID, precision, and recall for CelebA and CIFAR-10.
Systematic review of 145 papers on LLM-based log analysis, providing a unified taxonomy, common design patterns, evaluation practices, and challenges for deployment under drift and limited labels.
MACDAE infers implicit contexts via a constrained autoencoder and integrates them into an end-to-end O2O recommender, reporting gains on Yelp/Dianping/Koubei and 2.9%/5.6% lifts in online CTR/conversion.
Denoising autoencoder pretraining on corrupted visual embeddings yields more robust Med-VQA performance on SLAKE and PathVQA while using LoRA for efficient LLM adaptation.
citing papers explorer
-
MeshTok: Efficient Multi-Scale Tokenization for Scalable PDE Transformers
MeshTok uses AMR-inspired adaptive multiscale tokenization to improve the efficiency-accuracy trade-off of Transformer models for PDEs over uniform-grid baselines.
-
LLM-guided Semi-Supervised Approaches for Social Media Crisis Data Classification
LG-CoTrain, an LLM-guided co-training method, outperforms classical semi-supervised baselines for crisis tweet classification in low-resource settings with 5-25 labeled examples per class.
-
Explaining the effects of non-convergent sampling in the training of Energy-Based Models
EBMs trained with non-persistent short runs reproduce empirical data statistics via a precise dynamical process, not the equilibrium measure.
-
A Geometric Lens on Physics-Aligned Data Compression
Develops a local tangent-space rate-distortion theory and eigenspace-overlap diagnostic showing when physics-aligned compression necessarily degrades standard fidelity due to misaligned sensitivity directions.
-
Energy-based models for diagnostic reconstruction and analysis in a laboratory plasma device
A single energy-based model trained on LAPD plasma data enables diagnostic reconstruction, inverse inference of probe position, conditional trend sampling, and unconditional mode reproduction for potential anomaly detection.
-
Physics-Informed Tracking (PIT)
PIT uses a neural autoencoder with a differentiable physics module and a new Physics-Informed Landmark Loss to track single particles in video, achieving sub-pixel accuracy in supervised and unsupervised modes.
-
Random Vector Functional Link Neural Network based Ensemble Deep Learning
The authors introduce dRVFL and edRVFL frameworks that stack RVFL layers with fixed random weights and closed-form outputs, reporting superior benchmark performance when combined with sparse-pretrained RVFL.
-
Multi-Block Attention for Efficient Channel Estimation in IRS-Assisted mmWave MIMO
A multi-block attention neural network reduces pilot overhead by 87% and NMSE by 51% at 10 dB SNR for cascaded channel estimation in IRS-assisted mmWave MIMO-OFDM systems.
-
PixelFlowCast: Latent-Free Precipitation Nowcasting via Pixel Mean Flows
PixelFlowCast delivers high-fidelity precipitation nowcasts from radar sequences using a latent-free Pixel Mean Flows predictor guided by a deterministic coarse stage and KANCondNet features.
-
On the Redundancy of Timestep Embeddings in Diffusion Models
Timestep embeddings are redundant in diffusion models under certain conditions, with time-agnostic variants matching or exceeding conditioned models on FID, precision, and recall for CelebA and CIFAR-10.
-
LLM4Log: A Systematic Review of Large Language Model-based Log Analysis
Systematic review of 145 papers on LLM-based log analysis, providing a unified taxonomy, common design patterns, evaluation practices, and challenges for deployment under drift and limited labels.
-
Infer Implicit Contexts in Real-time Online-to-Offline Recommendation
MACDAE infers implicit contexts via a constrained autoencoder and integrates them into an end-to-end O2O recommender, reporting gains on Yelp/Dianping/Koubei and 2.9%/5.6% lifts in online CTR/conversion.
-
Noise-Aware Visual Representation Learning for Medical Visual Question Answering
Denoising autoencoder pretraining on corrupted visual embeddings yields more robust Med-VQA performance on SLAKE and PathVQA while using LoRA for efficient LLM adaptation.