ARMA nets expand receptive fields for dense prediction tasks.
problem Global information in dense prediction problems is challenging for traditional convolutional layers.
method ARMA layers with adjustable autoregressive coefficients replace traditional convolutions.
result ARMA networks improve dense prediction tasks including video prediction and semantic segmentation.
A new method compresses deep neural networks by predicting and quantizing weights between layers.
problem Resource constraints in deep neural networks.
method Inter-Layer Weight Prediction (ILWP) and quantization based on Smoothly Varying Weight Hypothesis (SVWH).
result The method achieves higher weight compression rates at the same accuracy level.
Proposes LML layer for multi-label predictions with k labels.
problem Efficient multi-label prediction with limited labels.
method Probabilistic multi-label modeling, efficient forward and backward passes.
result Improves top-k recall and accuracy in multi-label tasks.
DCFS predicts stock indices using deep learning and fuzzy systems.
problem Predicting stock indices with high-dimensional inputs.
method Bottom-up layer-by-layer design of DCFS, parameter sharing.
result DCFS models outperform weak estimators and real-world data.
New CNN approach reduces overconfidence in object classification predictions.
problem Overconfident predictions from deep models, especially SoftMax layer.
method Introduces CNN probabilistic approach using Logit layer for Bayesian inference.
result Proposed approach shows promising performance compared to SoftMax.
A deep learning network was used to predict future blood glucose levels, as this can permit diabetes patients to take action before imminent hyperglycaemia and hypoglycaemia. A sequential model with one long-short-term memory (LSTM) layer, one bidirectional LSTM layer and several fully connected layers was used to pred…
StackNet predicts fluid intelligence from brain images of adolescents.
problem Predicting fluid intelligence in adolescents using brain imaging.
method Feature extraction, normalization, denoising, selection, StackNet architecture, 11 models, 3 layers, 10-fold cross-validation.
result StackNet achieves mean squared errors of 82.42 on training/validation and 94.25 on testing.
A method predicts GNS of transformer layers using normalization layer norms.
problem Estimating gradient noise scale with minimal variance.
method Simultaneously compute per-example gradient norms and parameter gradients.
result Total GNS is predicted well by normalization layer GNS.
Paper proposes a method to predict disk failures using multi-layer domain adaptive learning.
problem Traditional machine learning models struggle to predict disk failures due to limited data.
method Multi-layer domain adaptive learning with source and target domains.
result The proposed method improves failure prediction accuracy on disk data with few failure samples.
Stacked LSTM improves weather forecasting accuracy by incorporating spatial information.
problem Improving temperature prediction accuracy in weather forecasting.
method 2-layer spatio-temporal stacked LSTM model with independent LSTM models per location in the first layer and combined hidden states in the second layer.
result The stacked LSTM model outperforms single LSTM models in most cases by utilizing spatial information.
Deep learning uses layers of transformations to predict structured data with uncertainty.
problem Predicting structured high-dimensional data efficiently and with uncertainty.
method Applying layers of semi-affine input transformations to find features for probabilistic statistical methods.
result Achieves scalable prediction rules with uncertainty quantification and feature selection.
Optimizes CNN architectures by analyzing receptive fields without training.
problem CNNs often have too many layers, making them resource-intensive and less effective.
method Layer-wise analysis of receptive fields to identify unproductive layers.
result Identifies and removes unproductive layers, improving CNN performance and efficiency.
Many real-world complex systems are well represented as multilayer networks; predicting interactions in those systems is one of the most pressing problems in predictive network science. To address this challenge, we introduce two stochastic block models for multilayer and temporal networks; one of them uses nodes as it…
Algorithm predicts performance of learning in multi-layer networks with matrix-valued hidden variables.
problem Signal recovery and learning in multi-layer neural networks with matrix-valued hidden variables.
method Unified approximation algorithm for MAP and MMSE inference, extending ML-VAMP to handle matrix-valued unknowns.
result Performance of ML-Mat-VAMP algorithm can be predicted in a random large-system limit.
A new metric measures saturation of neural network layers.
problem Analyzing the quality of latent representations in neural networks.
method Layer Saturation metric based on spectral analysis.
result Saturation is related to generalization and predictive performance.
CNNs encode data augmentation transformations, especially in early layers.
problem Whether neural network features encode data augmentation transformations.
method Systematic approach using pre-trained vision models to predict augmentation transformations.
result Neural network features encode data augmentation transformations, especially in early layers.
Introduces RPL for improved class prediction and uncertainty assessment.
problem Softmax's limitations in calculating class probabilities and handling uncertainty.
method Develops a new prediction layer (RPL) based on the open-world assumption.
result RPL provides more meaningful class prediction probabilities and handles uncertainty better.
Adds layers to NNs to protect them from reverse engineering.
problem Extracting the underlying model of a Neural Network.
method Introducing parasitic layers that approximate a noisy identity mapping with a Convolutional NN.
result The protected NN's predictions remain mostly unchanged while making reverse-engineering more complex.
New approach optimizes sales process for B2B businesses.
problem Optimizing the sales process for B2B businesses.
method Causal Predictive Optimization and Generation with three layers: prediction, optimization, and serving.
result Significant wins over legacy systems in LinkedIn implementation.
This study shows unstructured clinical notes can improve mortality prediction.
problem Lack of effective use of unstructured clinical notes in mortality prediction.
method Used a hierarchical architecture with convolutional and recurrent layers to predict in-hospital mortality from unprocessed clinical notes.
result Achieved higher metrics in mortality prediction compared to structured data approaches.
DeepLight accelerates CTR predictions in ad serving by 46X.
problem Significantly increased serving delay and high memory usage for ad serving.
method Explicitly searching feature interactions, pruning layers, promoting sparsity.
result Accelerates model inference by 46X on Criteo dataset.
Predictive Sparse Manifold Transform learns dynamic video sequences.
problem Learning and predicting natural dynamics in video sequences.
method Two-layer framework: sparse coding and manifold learning.
result PSMT with dynamic embedding space outperforms static baselines in future frame prediction.
SAEs struggle with curved activation manifolds, revealing layer-dependent scaling laws.
problem Sparse autoencoders' reconstruction error varies across layers, not fitting existing scaling laws.
method Cross-layer study of 844 SAE checkpoints, fitting and regressing on manifold geometry.
result Manifold geometry predicts layer-dependent width exponents in SAEs, with transferable coefficients.
Law explains how LLMs learn to predict next tokens.
problem Understanding how LLMs process input data internally.
method Introduced a precise law governing token embeddings in LLMs.
result Each layer equally contributes to next-token prediction accuracy.
Develops LSTM for predicting neuronal dynamics over long time-horizons.
problem Understanding and controlling complex brain behaviors.
method Long Short-Term Memory (LSTM) neural network architecture for multi-time step predictions.
result LSTM improves short time-horizon prediction accuracy and multi-time step predictions of neuronal dynamics.
The fully connected layers of a deep convolutional neural network typically contain over 90% of the network parameters, and consume the majority of the memory required to store the network parameters. Reducing the number of parameters while preserving essentially the same predictive performance is critically important …
Motivation: Understanding functions of proteins in specific human tissues is essential for insights into disease diagnostics and therapeutics, yet prediction of tissue-specific cellular function remains a critical challenge for biomedicine. Results: Here we present OhmNet, a hierarchy-aware unsupervised node feature le…
Measures difficulty of predictions to improve deep learning models.
problem Understanding and improving deep learning models.
method Introduces a measure called prediction depth to assess example difficulty.
result Surprising relationships between prediction depth and model performance.
Integrates fairness guarantees into deep learning models.
problem Ensuring fairness in deep learning models.
method Integrates a differentiable fairness layer into neural models and uses an online primal-dual algorithm for provable fairness guarantees.
result Guarantees a chosen notion of output parity in deep learning models.
Proposes a new criterion for reliable uncertainty estimation in deep neural networks.
problem Inability of existing approaches to provide reliable uncertainty estimates for deep neural networks.
method Develops a density uncertainty layer architecture that satisfies the proposed criterion.
result Density uncertainty layers provide more reliable uncertainty estimates and robust out-of-distribution detection.
Looped transformers with LN converge to power method for principal component prediction.
problem Understanding how transformers learn algorithmic procedures.
method Study of principal component prediction with looped linear transformers and layer normalization.
result Gradient descent trains looped transformers with LN to implement the power method for principal component prediction.
Paper speeds up neural language model inference by 20x for top-k word prediction.
problem Slow inference speed of neural language models on mobile devices.
method Introduced a screening model using Gumbel softmax to approximate softmax layer.
result Achieved 20.4x speedup with 98.9% precision@1 and 99.3% precision@5 for German to English translation.
The paper analyzes the training dynamics of a transformer for next-token prediction.
problem Understanding the non-asymptotic performance of transformers in next-token prediction.
method Characterizes training dataset properties, designs a two-stage training algorithm, and analyzes attention gradient properties.
result Trained transformers converge sub-linearly to max-margin solutions and exhibit linear convergence in cross-entropy loss.
Bayesian neural networks reveal multimodal predictive distributions.
problem Uncertainty quantification and interpretability in neural networks.
method Discretized prior for inner layer weights, Gaussian mixture approximation of posterior predictive distribution.
result Distinct parameter realizations can produce the same training error but different posterior predictive distributions.
SmallML predicts customer churn for SMEs with small data, improving accuracy by 24.2 points.
problem AI exclusion of SMEs due to data scale mismatch.
method Bayesian transfer learning with hierarchical pooling and conformal prediction.
result 96.7% AUC on 100 obs SMEs, 24.2 point improvement over logistic regression.
Differentiable Masking reveals how neural models make decisions across layers.
problem Intractable and expensive approximate search for input relevance in deep models.
method Differentiable Masking learns to mask inputs while maintaining differentiability.
result Reveals how decisions are formed across network layers in BERT models.
New method improves BLL models for complex datasets.
problem Limited expressive capacity of Gaussian priors in BLL models.
method Combines diffusion techniques and implicit priors for variational learning.
result Enhanced predictive accuracy and uncertainty quantification.
A new readout layer improves graph property prediction.
problem Graph property prediction challenges due to information loss.
method Introduces a novel generalized global pooling layer for Message-Passing Neural Networks.
result New state-of-the-art results in graph property prediction.
Axial-LOB predicts stock prices from LOB data using attention layers.
problem Predicting stock price from LOB data with long-range dependencies.
method Axial-LOB uses gated position-sensitive axial attention layers to incorporate global interactions.
result Axial-LOB achieves state-of-the-art performance in stock price prediction.
The introduction of convolutional layers greatly advanced the performance of neural networks on image tasks due to innately capturing a way of encoding and learning translation-invariant operations, matching one of the underlying symmetries of the image domain. In comparison, there are a number of problems in which the…
LoRA fine-tuning creates intruder dimensions that can cause forgetting, and a new law predicts when this happens.
problem Predicting when LoRA fine-tuning creates intruder dimensions that can cause catastrophic forgetting.
method Derived a per-layer critical update strength s∗ and an exact secular-equation characterization of the updated spectrum. result The law localizes the empirical threshold within a factor of two on 82% of layers and separates intruder-bearing from intruder-free layers at deployment.
EDAIN layer normalizes time series data for neural networks, improving model performance.
problem Irregularities in time series data degrade model performance in neural networks.
method EDAIN layer learns adaptive normalization parameters during end-to-end training.
result EDAIN layer outperforms conventional normalization methods and adaptive layers.
Deep neural network predicts traffic flow on city maps.
problem Short-term traffic flow prediction on high-resolution city maps.
method UNet-based deep convolutional neural network with densely connected layers.
result Best performance on the Traffic4cast challenge 2019.
Self-supervised learning improves by predicting known information, reducing labeled data needs.
problem Efficiently learn useful semantic representations without labeled data.
method Develops a mechanism exploiting statistical connections between pretext tasks to learn representations that solve downstream tasks.
result Proves linear layer yields small approximation error and drastically reduces labeled sample complexity.
A hybrid method combines GA and EN for feature selection in high-dimensional datasets.
problem Feature selection in high-dimensional datasets with high prediction error and computational inefficiency.
method Hybrid two-layer approach using Genetic Algorithm and Elastic Net.
result The hybrid method improves prediction accuracy and reduces computational time.
End-to-end network predicts and aligns continuous emotion labels.
problem Inconsistent alignment of continuous emotion labels with speech signals.
method Convolutional neural network with a multi-delay sinc layer.
result State-of-the-art results in predicting and aligning continuous emotion labels.
InteractionNet models noncovalent protein-ligand interactions with GNNs and explains predictions.
problem Modeling noncovalent protein-ligand interactions with graph neural networks.
method InteractionNet uses a GNN architecture with separated covalent and noncovalent convolution layers and layer-wise relevance propagation for explainability.
result InteractionNet successfully predicts noncovalent protein-ligand interactions with chemical relevance.
EagerNet detects network attacks quickly with less resources.
problem Efficiently detecting network attacks with minimal resources.
method Proposes a new architecture that trades prediction speed for accuracy, evaluating only a subset of layers.
result Comparable accuracies to simple FCNNs achieved with early predictions, saving energy and computational efforts.