A systematic approach maps any-dimensional invariant functions to a unique function on an infinite-dimensional limit space admitting a topology with compact sets where universality holds, with examples of non-universal architectures and fixes.
super hub Mixed citations
Attention is all you need.Advances in neural information processing systems, 30
Mixed citation behavior. Most common role is background (52%).
138
Pith papers citing it
Background
52% of classified citations
hub tools
citation-role summary
background 17
method 12
baseline 2
dataset 1
other 1
citation-polarity summary
claims ledger
- background and block-sparse FlashAttentionenable longer context in Transformers, yielding higher quality models (0.7 better perplexity on GPT-2 and 6.4 points of lift on long-document classification) and entirely new capabilities: the first Transformers to achieve better-than-chance performance on the Path-X challenge (seq. length 16K, 61.4% accuracy) and Path-256 (seq. length 64K, 63.1% accuracy). 1 Introduction Transformer models [82] have emerged as the most widely used architecture in applications such a
- method would differ in the network structure (i.e., G̸=G ′). Then, there exists (f, g) such that the population trajectories({x i(t+k)}, G(t+k)and({x ′ i(t+k)}, G ′(t+k)diverge for allk >0. Single-task agentic systems either treat observations as independent and identically distributed (i.i.d.) [93] or the dependencies are modeled globally through a full attention mechanism [ 95]. Neither captures the topology-constrained local observability that characterizes real social systems. In a MASS, G is an ir
authors
co-cited works
representative citing papers
New lower bounds establish that Deep Sets need embedding dimension linear in the number of points (up to constants) for d>1, and give the first non-trivial bounds for higher-order Janossy pooling.
EQ-VMamba adds rotation-equivariant cross-scan and group Mamba blocks to enforce end-to-end rotation equivariance, yielding better rotation robustness, competitive accuracy, and roughly 50% fewer parameters than non-equivariant baselines across classification, segmentation, and super-resolution.
k-WL is incomplete on simple spectrum graphs; PRiSM is the first provably complete canonicalization for their eigendecompositions.
CDM amortizes SMC inference for reward-tilted discrete diffusion by training a parameterized twist function on contrastive samples with closed-form kernels.
CRiSP uses neural-guided MCTS and curriculum learning to insert Clifford prefixes before parameterized rotations in VQAs, yielding mean 3.17x and max 45x gains in energy accuracy on 22-qubit QAOA benchmarks versus prior Clifford initializers.
FISolver trains a compact LLM on backward-generated (differential equation, first integral) pairs and uses guided reinforcement learning to outperform larger models and Mathematica on first-integral benchmarks at lower cost.
A hypernetwork generates complete task-specific visuomotor policy parameters from instructions alone to structurally eliminate observation leakage in language-conditioned robotic control.
Hybrid TimesFM plus ridge regression on covariates forecasts 1-MeV electron flux with average R² of 0.9 on out-of-sample 2024 data, outperforming linear regression, CNN, LSTM and Transformer models.
DCDM replaces positional blocks with learnable semantic chunks via differentiable Chunking Attention, yielding consistent gains over block and unstructured diffusion baselines up to 1.5B parameters.
SurvivalPFN amortizes Bayesian survival analysis for right-censored data by pretraining a prior-data fitted network on synthetic identifiable DGPs and then performing in-context inference, achieving competitive results on 61 real datasets.
Local neural operators on 3x3x3 patches, composed via Schwarz iteration, solve large-scale nonlinear elasticity on arbitrary geometries without domain-specific retraining.
Contrastive predictive coding pretraining combined with structured state space models yields the strongest ECG foundation models, with continued gains from scaling data to 11 million samples.
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.
DeepLévy learns mixtures of Lévy stable distributions for heavy-tailed time series forecasting by minimizing discrepancies between empirical and parametric characteristic functions, outperforming prior methods on tail risk metrics under extreme volatility.
TIDES reconciles selective SSM expressivity with continuous-time physical discretization by moving input dependence onto the state matrix, enabling native irregular time series handling and achieving SOTA on UEA and Physiome-ODE benchmarks.
LENS shapes low-frequency eigen noise with a lightweight network to enable efficient, high-quality sampling in distilled diffusion models.
S2M extracts structured text quadruples from change masks to provide noise-free multimodal supervision, achieving 17.80% Sek and 66.14% F_scd on the new Gaza-Change-v2 dataset and outperforming LLM-based multimodal methods.
LookWhen factorizes video recognition into learning when, where, and what to compute via uniqueness-based token selection and dual-teacher distillation, achieving better accuracy-FLOPs trade-offs than baselines on multiple datasets.
DAPRO provides the first dynamic, theoretically guaranteed way to allocate interaction budgets across test cases for bounding time-to-event in multi-turn LLM evaluations, achieving tighter coverage than static conformal survival methods.
FLUID is a continuous-time transformer using Liquid Attention Networks to model attention as stable ODE solutions that interpolate between discrete SDPA and CT-RNNs, with an explicit sink gate and liquid hyper-connections for better information flow.
TrajShield is a training-free defense that reduces jailbreak success rates by 52.44% on average in text-to-video models by localizing and neutralizing risks through trajectory simulation and causal intervention.
Evolutionary trees from LLM weights recover ground-truth training topologies and identify key datasets and layers through phenotypic analysis.
Stream-CQSA uses CQS-based decomposition to stream exact attention computations for billion-token sequences on limited-memory hardware.
citing papers explorer
-
Any-Dimensional Invariant Universality
A systematic approach maps any-dimensional invariant functions to a unique function on an infinite-dimensional limit space admitting a topology with compact sets where universality holds, with examples of non-universal architectures and fixes.
-
Embedding Dimension Lower Bounds for Universality of Deep Sets and Janossy Pooling
New lower bounds establish that Deep Sets need embedding dimension linear in the number of points (up to constants) for d>1, and give the first non-trivial bounds for higher-order Janossy pooling.
-
Rotation Equivariant Mamba for Vision Tasks
EQ-VMamba adds rotation-equivariant cross-scan and group Mamba blocks to enforce end-to-end rotation equivariance, yielding better rotation robustness, competitive accuracy, and roughly 50% fewer parameters than non-equivariant baselines across classification, segmentation, and super-resolution.
-
Weisfeiler-Leman Is Incomplete on Simple Spectrum Graphs, so Canonicalize Them
k-WL is incomplete on simple spectrum graphs; PRiSM is the first provably complete canonicalization for their eigendecompositions.
-
Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusion
CDM amortizes SMC inference for reward-tilted discrete diffusion by training a parameterized twist function on contrastive samples with closed-form kernels.
-
Classical State Preparation for Variational Quantum Algorithms via Reinforcement Learning
CRiSP uses neural-guided MCTS and curriculum learning to insert Clifford prefixes before parameterized rotations in VQAs, yielding mean 3.17x and max 45x gains in energy accuracy on 22-qubit QAOA benchmarks versus prior Clifford initializers.
-
Learning First Integrals via Backward-Generated Data and Guided Reinforcement Learning
FISolver trains a compact LLM on backward-generated (differential equation, first integral) pairs and uses guided reinforcement learning to outperform larger models and Mathematica on first-integral benchmarks at lower cost.
-
DISC: Decoupling Instruction from State-Conditioned Control via Policy Generation
A hypernetwork generates complete task-specific visuomotor policy parameters from instructions alone to structurally eliminate observation leakage in language-conditioned robotic control.
-
Forecasting megaelectron-volt electron flux in the Earth's outer radiation belt using supervised machine learning algorithms and a timeseries foundation model
Hybrid TimesFM plus ridge regression on covariates forecasts 1-MeV electron flux with average R² of 0.9 on out-of-sample 2024 data, outperforming linear regression, CNN, LSTM and Transformer models.
-
Dynamic Chunking for Diffusion Language Models
DCDM replaces positional blocks with learnable semantic chunks via differentiable Chunking Attention, yielding consistent gains over block and unstructured diffusion baselines up to 1.5B parameters.
-
SurvivalPFN: Amortizing Survival Prediction via In-Context Bayesian Inference
SurvivalPFN amortizes Bayesian survival analysis for right-censored data by pretraining a prior-data fitted network on synthetic identifiable DGPs and then performing in-context inference, achieving competitive results on 61 real datasets.
-
Neural-Schwarz Tiling for Geometry-Universal PDE Solving at Scale
Local neural operators on 3x3x3 patches, composed via Schwarz iteration, solve large-scale nonlinear elasticity on arbitrary geometries without domain-specific retraining.
-
Pretraining Strategies and Scaling for ECG Foundation Models: A Systematic Study
Contrastive predictive coding pretraining combined with structured state space models yields the strongest ECG foundation models, with continued gains from scaling data to 11 million samples.
-
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.
-
DeepL\'evy: Learning Heavy-Tailed Uncertainty in Highly Volatile Time Series
DeepLévy learns mixtures of Lévy stable distributions for heavy-tailed time series forecasting by minimizing discrepancies between empirical and parametric characteristic functions, outperforming prior methods on tail risk metrics under extreme volatility.
-
TIDES: Implicit Time-Awareness in Selective State Space Models
TIDES reconciles selective SSM expressivity with continuous-time physical discretization by moving input dependence onto the state matrix, enabling native irregular time series handling and achieving SOTA on UEA and Physiome-ODE benchmarks.
-
LENS: Low-Frequency Eigen Noise Shaping for Efficient Diffusion Sampling
LENS shapes low-frequency eigen noise with a lightweight network to enable efficient, high-quality sampling in distilled diffusion models.
-
Masks Can Talk: Extracting Structured Text Information from Single-Modal Images for Remote Sensing Change Detection
S2M extracts structured text quadruples from change masks to provide noise-free multimodal supervision, achieving 17.80% Sek and 66.14% F_scd on the new Gaza-Change-v2 dataset and outperforming LLM-based multimodal methods.
-
LookWhen? Fast Video Recognition by Learning When, Where, and What to Compute
LookWhen factorizes video recognition into learning when, where, and what to compute via uniqueness-based token selection and dual-teacher distillation, achieving better accuracy-FLOPs trade-offs than baselines on multiple datasets.
-
How Many Iterations to Jailbreak? Dynamic Budget Allocation for Multi-Turn LLM Evaluation
DAPRO provides the first dynamic, theoretically guaranteed way to allocate interaction budgets across test cases for bounding time-to-event in multi-turn LLM evaluations, achieving tighter coverage than static conformal survival methods.
-
FLUID: Continuous-Time Hyperconnected Sparse Transformer for Sink-Free Learning
FLUID is a continuous-time transformer using Liquid Attention Networks to model attention as stable ODE solutions that interpolate between discrete SDPA and CT-RNNs, with an explicit sink gate and liquid hyper-connections for better information flow.
-
TrajShield: Trajectory-Level Safety Mediation for Defending Text-to-Video Models Against Jailbreak Attacks
TrajShield is a training-free defense that reduces jailbreak success rates by 52.44% on average in text-to-video models by localizing and neutralizing risks through trajectory simulation and causal intervention.
-
Analysis and Explainability of LLMs Via Evolutionary Methods
Evolutionary trees from LLM weights recover ground-truth training topologies and identify key datasets and layers through phenotypic analysis.
-
Stream-CQSA: Avoiding Out-of-Memory in Attention Computation via Flexible Workload Scheduling
Stream-CQSA uses CQS-based decomposition to stream exact attention computations for billion-token sequences on limited-memory hardware.
-
A Hormone-inspired Emotion Layer for Transformer language models (HELT)
HormoneT5 augments T5 with a hormone-inspired block that predicts six continuous emotion values and uses them to modulate responses, reporting over 85% per-hormone accuracy and human preference for emotional quality.
-
Characterizing Performance-Energy Trade-offs of Large Language Models in Multi-Request Workflows
This work delivers the first measurements of performance-energy trade-offs across four multi-request LLM workflow patterns on A100 GPUs using vLLM and Parrot.
-
Turing Test on Screen: A Benchmark for Mobile GUI Agent Humanization
The work creates a new benchmark for humanizing GUI agent touch dynamics via a MinMax detector-agent model, a mobile touch dataset, and methods showing agents can match human behavior without losing task performance.
-
BadSNN: Backdoor Attacks on Spiking Neural Networks via Adversarial Spiking Neuron
BadSNN injects backdoors into spiking neural networks by adversarially tuning LIF neuron hyperparameters and optimizing triggers, achieving higher attack success than prior data-poisoning methods while remaining robust to common defenses.
-
ModeX: Evaluator-Free Best-of-N Selection for Open-Ended Generation
ModeX selects the modal semantic output from multiple LLM generations via a similarity graph and recursive spectral clustering without needing reward models or evaluators.
-
Cognitive Alpha Mining via LLM-Driven Code-Based Evolution
CogAlpha combines LLM reasoning with code-level evolutionary search to discover financial alphas that show higher predictive accuracy and generalization than prior methods on five stock datasets.
-
Token Buncher: Shielding LLMs from Harmful Reinforcement Learning Fine-Tuning
TokenBuncher constrains response entropy via entropy-as-reward RL and a Token Noiser to stop harmful RL fine-tuning while keeping benign performance intact.
-
CAD-Coder: Text-to-CAD Generation with Chain-of-Thought and Geometric Reward
CAD-Coder generates valid CadQuery scripts from text via supervised fine-tuning followed by reinforcement learning with geometric Chamfer Distance rewards and chain-of-thought planning.
-
Smoothie: Smoothing Diffusion on Token Embeddings for Text Generation
Smoothie performs diffusion by smoothing token embeddings based on semantic similarity, outperforming prior diffusion models on sequence-to-sequence and unconditional text generation tasks.
-
FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precision
FlashAttention-3 achieves 1.5-2x speedup on H100 GPUs for attention, reaching 740 TFLOPs/s (75% utilization) in FP16 and near 1.2 PFLOPs/s in FP8 while cutting numerical error by 2.6x versus baseline FP8 attention.
-
LAION-5B: An open large-scale dataset for training next generation image-text models
LAION-5B is an openly released dataset of 5.85 billion CLIP-filtered image-text pairs that enables replication of foundational vision-language models.
-
FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness
FlashAttention reduces GPU high-bandwidth memory accesses in self-attention via tiling, delivering exact attention with lower IO complexity, 2-3x wall-clock speedups on models like GPT-2, and the ability to train on sequences up to 64K long.
-
Winner-Take-All bottlenecks enforce disentangled symbolic representations in multi-task learning
WTA bottlenecks enforce highly symbolic, disentangled categorical representations of latent factors under defined conditions in multi-task DNNs, shown via theorem and experiments on two datasets.
-
Distributional Alignment as a Criterion for Designing Task Vectors in In-Context Learning
A distributional alignment metric d_NTP and a linear regression method LTV for task vectors that improves accuracy by 9.2% over baselines on classification and regression tasks across multiple LLMs.
-
Accelerating Video Inverse Problem Solvers with Autoregressive Diffusion Models
AVIS applies autoregressive diffusion models to video inverse problems by streaming restoration with measurement-consistent initialization, reducing latency from 114s to 4s and raising throughput to 1.18 FPS (or 5.91 FPS in the Flash variant).
-
TrajTok: Adaptive Spatial Tokenization for Trajectory Representation Learning
TrajTok learns multi-resolution hexagonal spatial tokens from GPS data and pretrains a factorized transformer with ST-RoPE and masked modeling to yield frozen encoders that outperform task-specific methods on similarity, classification, and travel-time tasks in the Porto dataset.
-
Generative Recursive Reasoning
GRAM is a latent-variable generative model that performs recursive reasoning via stochastic trajectories, trained with amortized variational inference to support multi-hypothesis reasoning and unconditional generation.
-
DashAttention: Differentiable and Adaptive Sparse Hierarchical Attention
DashAttention introduces differentiable adaptive sparse hierarchical attention via α-entmax block selection, achieving full-attention accuracy at 75% sparsity with improved Pareto performance over NSA and InfLLMv2.
-
PIXLRelight: Controllable Relighting via Intrinsic Conditioning
A transformer-based neural renderer that transfers arbitrary PBR lighting to single images via shared intrinsic conditioning extracted from both multi-illumination photos and path-traced coarse 3D renders.
-
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.
-
Registers Matter for Pixel-Space Diffusion Transformers
Register tokens enhance pixel-space DiT training and output quality via cleaner high-noise feature maps, and a dual-stream design adds further gains with little overhead.
-
Invaria: Learning Scale and Density Invariance in Point Clouds via Next-Resolution Prediction
Invaria trains point cloud encoders with next-resolution prediction to learn scale and density invariant features, yielding higher mIoU on ScanNet under lower resolution and scaled objects while using a smaller model.
-
Nonlinear Bipolar Compensation: Handling Outliers in Post-Training Quantization
Nonlinear Bipolar Compensation with Bipolar Logarithmic Transformation reduces outlier effects in post-training quantization by performing compensation in a compressed transformed space.
-
OmniDrop: Layer-wise Token Pruning for Omni-modal LLMs via Query-Guidance
OmniDrop is a training-free layer-wise token pruning framework for omni-modal LLMs that uses query guidance and temporal diversity to reduce prefill latency by up to 40% and memory by 14.7% while improving benchmark scores by up to 3.58 points.
-
InsightTok: Improving Text and Face Fidelity in Discrete Tokenization for Autoregressive Image Generation
InsightTok improves text and face fidelity in discrete image tokenization via content-aware perceptual losses, with gains transferring to autoregressive generation.
-
CO-MAP: A Reinforcement Learning Approach to the Qubit Allocation Problem
Reinforcement learning policy for qubit mapping reduces SWAP overhead by 65-85% versus standard quantum compilers on MQTBench and Queko benchmark circuits.