New framework shows cross-attention improves multi-modal in-context learning.
problem Understanding multi-modal in-context learning in neural networks.
method Mathematical framework and linearized cross-attention mechanism.
result Cross-attention mechanism is provably optimal for multi-modal in-context learning.
Generative models use kernel smoothing for conditioning on small example sets.
problem Improving generative models' performance with limited conditioning examples.
method Showed that cross-attention conditioning is equivalent to kernel smoothing, specifically a Nadaraya--Watson kernel smoother.
result The approach predicts and confirms three failure regimes for kernel-based conditioning.
DAGR improves navigation by refining goal representations conditioned on the current state.
problem Goal-conditioned reinforcement learning lacks state awareness, leading to inefficient policy recovery.
method DAGR refines static goal embeddings into state-conditioned ones using gated cross-attention with a state-goal discrepancy map.
result DAGR improves navigation tasks on OGBench, matching or outperforming base methods.
Improved GRU model with multi-head cross-attention enhances stock prediction accuracy.
problem Inaccurate stock prediction due to complex market dynamics and data sparsity.
method Enhanced GRU with multi-head cross-attention for better historical information selection and latent market state learning.
result The proposed MCI-GRU model outperforms state-of-the-art techniques in multiple metrics.
Single-head attention approximates any function under various norms.
problem Universal approximation of functions using attention mechanisms.
method Interpreting attention as partitioning and summing linear transformations.
result Single-head attention can approximate any continuous function under L∞-norm and Lebesgue integrable functions under Lp-norm. Conformer encoder reverses sequence in time dimension, affecting decoder training.
problem Reversal of sequence in Conformer encoder impacts decoder training.
method Analyzed initial behavior of decoder cross-attention and proposed methods to avoid flipping.
result Self-attention module of Conformer starts dominating, allowing only reversed information to pass.
Study improves stock movement prediction using multimodal data.
problem Inaccurate stock movement prediction due to incomplete multimodal data integration.
method Introduces MSGCA framework for robust multimodal fusion.
result MSGCA framework outperforms existing methods by 21.7% on multimodal datasets.
AutoAnchor uses cross-attention to improve text-to-image model unlearning.
problem Mitigating harmful or copyrighted content in text-to-image models.
method Two-stage framework that automatically synthesizes manifold-proximal anchors using cross-attention consistency loss.
result Effective robust and unbiased unlearning across various baselines.
Recent state-of-the-art natural language understanding models, such as BERT and XLNet, score a pair of sentences (A and B) using multiple cross-attention operations - a process in which each word in sentence A attends to all words in sentence B and vice versa. As a result, computing the similarity between a query sente…
New model recognizes emotions with missing modalities, improving accuracy.
problem Handling missing modalities in emotion recognition.
method Transformer-based architecture with cross-attention and self-attention mechanisms.
result Improvement of 37% in predicting arousal values and 30% in valence values compared to baseline.
X-Trend quickly adapts to new financial regimes, increasing Sharpe ratio by 18.9%.
problem Adapting to rapidly changing financial market conditions.
method Few-shot learning and cross-attention mechanism.
result X-Trend increases Sharpe ratio by 18.9% over a neural forecaster and 10-fold over a conventional strategy.
Abstractor enhances Transformers for relational reasoning, improving sample efficiency and performance.
problem Improving sample efficiency and performance in relational tasks.
method Introduces Abstractor module with relational cross-attention to enable explicit relational reasoning.
result Dramatic improvements in sample efficiency and performance on various relational tasks.
RB-Modulation trains free diffusion models without external adapters.
problem Training-free personalization of diffusion models with style and content control.
method Stochastic optimal control with a style descriptor and cross-attention aggregation.
result Precise content and style extraction and control without external adapters.
Edge devices adapt pre-trained models to local data without backpropagation.
problem Adapting pre-trained models to edge devices' local data distributions.
method Feed-forward latent domain adaptation using cross-attention.
result Consistent improvements over ERM baselines and domain-supervised adaptation.
Fast Bayesian inference with adaptable priors for real-time applications.
problem Intractable exact posterior computation limits Bayesian inference's adoption.
method Distribution Transformer architecture that learns mappings between priors and posteriors.
result Significant reduction in computation time from minutes to milliseconds.
New method learns operators with geometric singularities from few samples.
problem Learning operators with geometric singularities from limited data.
method Double fibration transforms and cross-attention architectures.
result Operators can be learned superalgebraically from few samples.
Improved image synthesis with user scribbles and text prompts.
problem Inadequate details in generated images due to domain shift.
method Optimization problem formulation and cross-attention for control.
result Significant improvement in user satisfaction (85.32% higher).
Paper analyzes deep learning models for credit rating prediction using text and numerical data.
problem Improving credit rating prediction using multi-modal deep learning.
method Testing different deep learning models and fusion strategies for structured and unstructured datasets.
result CNN-based multi-modal model with two fusion strategies outperformed other models.
DeepDIVE disentangles input into marginal and conditional distributions for multi-task learning.
problem Challenges in multi-task learning due to conflicting objectives.
method Inspired by probability theory, DeepDIVE uses a variational autoencoder with disentangled features and cross-attention mechanism.
result DeepDIVE disentangles input and improves forecast accuracy compared to baseline models.
Framework tackles OOD challenges in molecule property prediction by modeling environments.
problem Challenges in modeling OOD samples for molecule property prediction.
method Soft causal learning framework incorporating chemistry theories and cross-attention mechanisms.
result Demonstrates well generalization ability on seven datasets.
A new method routes EEG covariance matrices across domains using adaptive subspace selection.
problem Challenges in cross-domain EEG decoding due to distinct SPD manifold regions.
method Dynamic Stiefel routing with expert filters and cross-attention for adaptive subspace projection.
result Consistent gains across three datasets: balanced accuracy improves from 0.773 to 0.823, 0.757 to 0.809, and 0.801 to 0.839.
Understanding the 3-dimensional structure of the world is a core challenge in computer vision and robotics. Neural rendering approaches learn an implicit 3D model by predicting what a camera would see from an arbitrary viewpoint. We extend existing neural rendering to more complex, higher dimensional scenes than previo…
New model uses attention for in-context learning of categorical data.
problem Learning from categorical data in context.
method Attention-based network with self-attention and cross-attention layers, using functional gradient descent.
result Model can perform multi-step inference for categorical observations.
Self attention mechanisms have become a key building block in many state-of-the-art language understanding models. In this paper, we show that the self attention operator can be formulated in terms of 1x1 convolution operations. Following this observation, we propose several novel operators: First, we introduce a 2D ve…
TAP transfers knowledge from unlabeled data to improve cross-modal learning.
problem Improving supervised learning performance using unlabeled data from a different modality.
method Probabilistic approach for missing information estimation, kernel regression, cross-attention module, TAP neural network.
result TAP significantly improves generalization across different domains and neural network architectures.
CT improves neural network performance on cell complex data.
problem Improving predictive performance of neural networks on complex data.
method Introducing the Cellular Transformer (CT) that generalizes graph-based transformers to cell complexes.
result CT achieves state-of-the-art performance on cell complex datasets without complex enhancements.
OTSeg uses multi-prompt Sinkhorn attention to improve zero-shot semantic segmentation.
problem Leveraging pre-trained CLIP knowledge to align text embeddings with pixel embeddings.
method OTSeg employs Multi-Prompts Sinkhorn (MPS) and Multi-Prompts Sinkhorn Attention (MPSA) to enhance semantic feature matching.
result OTSeg achieves state-of-the-art performance in zero-shot semantic segmentation tasks.
This paper predicts stock prices using LLMs and news embeddings.
problem Predicting stock prices with high accuracy and relevance.
method Integrates LLMs with stock name embeddings and attention mechanisms for news filtering.
result Reduces MAE by 7.11% compared to baseline.
Task oriented dialog agents provide a natural language interface for users to complete their goal. Dialog State Tracking (DST), which is often a core component of these systems, tracks the system's understanding of the user's goal throughout the conversation. To enable accurate multi-domain DST, the model needs to enco…
New model generates realistic single-cell gene expression data.
problem Generating realistic single-cell gene expression profiles is challenging.
method scLDM, a latent diffusion model using Diffusion Transformers and linear interpolants.
result Superior performance in generating realistic single-cell gene expression data.
CoFinDiff generates synthetic financial data capturing stylized facts and meeting specified conditions.
problem Limited data availability and difficulty in controlling synthetic financial data generation.
method Conditional diffusion model with cross-attention to incorporate conditions derived from price data.
result Synthetic data generated by CoFinDiff accurately meets specified conditions for trends and volatility.
CViT learns complex physical systems using vision transformer techniques.
problem Learning maps between infinite-dimensional function spaces in scientific machine learning.
method Combines vision transformer encoder, grid-based coordinate embedding, and cross-attention mechanism.
result Achieves state-of-the-art performance on multiple benchmarks, often surpassing larger models.
Develops a Causal Transformer for estimating counterfactual outcomes from longitudinal data.
problem Estimating counterfactual outcomes over time from observational data is challenging due to complex, long-range dependencies.
method Combines three transformer subnetworks with separate inputs for time-varying covariates, previous treatments, and previous outcomes into a joint network with in-between cross-attentions. Uses a custom, end-to-end training procedure with a counterfactual domain confusion loss to address confounding bias.
result Achieves superior performance over current baselines in synthetic and real-world datasets.
Improved multi-modal emotion recognition using deep learning.
problem Combining acoustic and text modalities for emotion recognition.
method Proposes a deep learning-based approach to fuse text and acoustic data using SincNet for acoustic features and parallel DCNN and Bi-RNN branches for text processing with cross attention.
result Achieves 3.5% improvement in weighted accuracy compared to existing methods.
Ground-A-Video edits videos without training, preserving intended changes.
problem Complex multi-attribute video editing with omitted or wrong changes.
method Grounding-guided video-to-video translation with Cross-Frame Gated Attention.
result Zero-shot multi-attribute video editing with improved accuracy and frame consistency.
We consider the large-scale query-document retrieval problem: given a query (e.g., a question), return the set of relevant documents (e.g., paragraphs containing the answer) from a large document corpus. This problem is often solved in two steps. The retrieval phase first reduces the solution space, returning a subset …
GyroSwin models plasma turbulence with neural nets, reducing costs and capturing neglected nonlinearities.
problem Understanding plasma turbulence in fusion reactors, which impairs confinement and limits reactor design.
method Introduces GyroSwin, a scalable 5D neural surrogate that approximates 5D nonlinear gyrokinetic simulations.
result GyroSwin outperforms reduced models in heat flux prediction and captures turbulent energy cascade.
MOCA uses modular attention to estimate causal effects from complex data.
problem Estimating causal effects from observational data with complex, non-linear, and high-dimensional treatment and outcome mechanisms.
method MOCA is a transformer-based framework that separates treatment and outcome modeling through modular design and one-way attention mechanism, with cutting-feedback to prevent outcome influence on treatment representations.
result MOCA outperforms classical estimators and machine learning approaches across various simulated and real-world scenarios.
GAttNHP predicts future events in temporal knowledge graphs by encoding long-range dependencies and handling mutual excitation.
problem Forecasting future events in temporal knowledge graphs due to long-range dependencies, mutual excitation, and heavy-tailed inter-arrival times.
method GAttNHP uses a self-attention encoder, semantic soft-grouping, and NCQ regression to address these issues.
result GAttNHP improves entity and time prediction on six benchmark TKG datasets compared to state-of-the-art baselines.
AaSP improves audio self-supervised learning by addressing aliasing issues.
problem Alias issues in audio spectrogram transformers.
method AaSP combines aliasing-aware patch representation, teacher-student masked modeling, cross-attention predictor, and contrastive regularization.
result AaSP learns more stable representations that integrate high-frequency cues.
HANET combines LSTM and attention mechanisms for better financial forecasting.
problem Lack of distinct macroeconomic regimes in financial datasets.
method Hierarchical Cross-Attention mechanism integrating long-run macro contexts with high-frequency market dynamics.
result HANET outperforms neural forecasters, especially during turbulent periods.
Temporal Causal Prior-Data Fitted Networks (TCPFN) for industrial time series causal discovery
problem Estimating causal effects in industrial time series
method Temporal Causal Prior-Data Fitted Networks
result Zero-shot causal discovery with explicit reliability signals