Few-shot learning with per-sample rich supervision reduces sample complexity.
problem Learning with few samples is challenging for deep models.
method Use per-sample semantic information to reduce sample complexity.
result Improved generalization error bound and online algorithm for few-shot learning.
TX-Ray analyzes and quantifies model knowledge transfer in NLP.
problem Insufficient methods for explaining and quantifying model knowledge transfer in NLP.
method Modified computer vision explainability principle to NLP, visualizing feature preference distributions.
result TX-Ray reveals how self-supervised models learn linguistic abstractions and improves generalization.
New method for differentially private optimization with general Lipschitz conditions.
problem Differentially private optimization under general Lipschitz conditions.
method Generalized Lipschitz condition for per-sample gradients, tuning clip norm based on minimum per-sample Lipschitz constant.
result Efficacy of the recommended clip norm tuning method verified on 8 datasets.
Generative models need per-sample confidence scores to improve quality and stability.
problem Generative models produce unreliable outputs and lack confidence measures.
method Flow Matching with Confidence (FMwC) injects noise and integrates it through the network, providing per-sample confidence scores.
result The confidence score correlates with the velocity field's divergence, offering insights into generative processes.
A new method improves training speed and robustness for deep learning models.
problem Training deep learning models with heavy-tailed gradient noise.
method Per-sample gradient clipping combined with SGD.
result PS-Clip-SGD achieves optimal convergence rates and high-probability guarantees.
Transformer model removes noise from light curves efficiently.
problem Challenges in processing astrophysical light curves due to noise.
method Denoising Time Series Transformer (DTST) model trained with masked objective.
result DTST model excels at removing noise and outliers in time series datasets.
A robust loss for anomaly mitigation and unsupervised contamination classification
problem Detecting and mitigating contamination in supervised and unsupervised settings
method Neural Bayesian Anomaly Mitigation (NBAM)
result Recovering the structure of contamination and identifying label-flip pairs
Develops method to train classifiers on incomplete feature datasets.
problem Training classifiers on datasets with missing feature subsets.
method Simultaneous training of neural networks and sparse coding.
result Classifier trained on incomplete features correctly separates original data.
New method assesses individual training points' privacy risk without retraining.
problem Privacy vulnerability of individual training points in membership inference attacks.
method Derives a closed-form decomposition of individual black-box MIA vulnerability, extending to deep networks.
result Proposes a surrogate score operating on last-layer representations that requires only a single trained model.
To solve key biomedical problems, experimentalists now routinely measure millions or billions of features (dimensions) per sample, with the hope that data science techniques will be able to build accurate data-driven inferences. Because sample sizes are typically orders of magnitude smaller than the dimensionality of t…
Interpretability has become an important topic of research as more machine learning (ML) models are deployed and widely used to make important decisions. Most of the current explanation methods provide explanations through feature importance scores, which identify features that are important for each individual input. …
This paper proposes a fast method for estimating input-dependent prediction intervals in Extreme Learning Machines.
problem Estimating reliable prediction intervals for Extreme Learning Machines with heteroscedastic outputs.
method A separate Extreme Learning Machine model estimates input-dependent prediction intervals using a weighted Jackknife method to correct for model uncertainty.
result The proposed method is fast, robust to heteroscedastic outputs, and handles large datasets and insufficient training data.
The parameters of temporal models, such as dynamic Bayesian networks, may be modelled in a Bayesian context as static or atemporal variables that influence transition probabilities at every time step. Particle filters fail for models that include such variables, while methods that use Gibbs sampling of parameter variab…
BatchTopK SAEs improve GPT-2 and Gemma activations with adjustable sparsity.
problem Interpreting language model activations using sparse autoencoders.
method Adapting the TopK constraint to the batch-level, allowing variable number of active latents per sample.
result BatchTopK SAEs consistently outperform TopK SAEs in reconstructing GPT-2 and Gemma activations.
New analysis shows how to balance privacy and accuracy in deep learning.
problem Balancing privacy and accuracy in deep learning models.
method Continuous time analysis through neural tangent kernel (NTK) for arbitrary architectures.
result Large clipping norm improves calibration without sacrificing accuracy.
Online TERM improves robustness and fairness in streaming data.
problem Streaming data's lack of worst-case fairness and robustness in ERM.
method Proposes an online TERM formulation to balance average-case accuracy with worst-case fairness and robustness.
result Negative tilting effectively suppresses outlier influence, positive tilting improves recall with minimal precision loss.
Paper proposes DP-SGD and DP-NSGD for differentially private non-convex optimization.
problem Mitigating privacy risks in large model learning.
method Clip or normalize per-sample gradients and add noise for differential privacy.
result Achieved convergence rate of gradient norm for non-convex optimization.
Enhanced reinforcement learning using ensemble methods.
problem Improving reinforcement learning performance.
method Distributional reinforcement learning with ensemble group-aided training.
result Ensemble methods lead to more robust and efficient learning.
New algorithm clusters LLM inputs efficiently with quality control.
problem Scaling LLM inference cost and latency for millions of users.
method Two-stage algorithm using Mini-batch K-Means and greedy selection of representatives.
result Enforces similarity and attribute guardrails exactly, scales to tens of millions of samples.
Algorithm estimates principal eigenvector with adaptive sensing, improving over non-adaptive methods.
problem Estimating principal eigenvector with limited scalar measurements.
method Compressed variant of Oja's algorithm using two adaptive measurements per sample.
result Convergence rate of O(λ1λ2d2/(Δ2t)) after t iterations, matching information-theoretic lower bound. We present a novel Metropolis-Hastings method for large datasets that uses small expected-size minibatches of data. Previous work on reducing the cost of Metropolis-Hastings tests yield variable data consumed per sample, with only constant factor reductions versus using the full dataset for each sample. Here we present…
New method enhances neural network robustness against adversarial attacks.
problem Enhancing neural network robustness against adversarial attacks.
method Variational framework with per-sample noise level selector.
result Enhanced empirical robustness and certified robustness.
Proposes a new method for uncertainty estimation in neural networks.
problem Uncertainty quantification in neural networks for high-risk applications.
method Intuitive framework based on signal-to-noise ratio and variance-gated measure.
result Demonstrates a collapse in diversity of committee machines.
We present a new data-driven benchmark system to evaluate the performance of new MCMC samplers. Taking inspiration from the COCO benchmark in optimization, we view this task as having critical importance to machine learning and statistics given the rate at which new samplers are proposed. The common hand-crafted exampl…
Clapping reduces memory usage in distributed optimization by reusing data samples.
problem Significant communication overhead and impractical memory overhead in pipeline-parallel distributed optimization.
method Lazy sampling strategy to reuse data samples across steps, supporting convergence without unbiased gradient assumptions.
result Clapping achieves convergence in few-epoch or online training regimes without sample-size memory overhead.
Method estimates differences between two functional graphs.
problem Estimating differences between two functional undirected graphical models.
method Direct estimation of graph difference, avoiding separate estimation.
result Method is consistent in high-dimensional settings.
This work tackles costly feature acquisition for classification, using RL.
problem Classification with costly feature acquisition.
method Formulated as a sequential decision-making problem (MDP) and solved with deep reinforcement learning.
result Robust performance across various datasets and settings, outperforming prior-art.
Traditional pairwise sequence alignment is based on matching individual samples from two sequences, under time monotonicity constraints. However, in many application settings matching subsequences (segments) instead of individual samples may bring in additional robustness to noise or local non-causal perturbations. Thi…
Develops an online learning framework for Bayesian joint filtering.
problem Streaming inference of nonlinear state-space models.
method Variational inference and sequential Monte Carlo.
result Efficient approximation of filtering posterior for a wide class of models.
Combines offline and online learning for identifying the best arm in bandits.
problem Identifying the best arm in stochastic bandits with offline data.
method Lower bound analysis and algorithm development for optimal best-arm identification.
result Developed algorithms matching lower bound on sample complexity for small δ.
Combining complementary information from multiple modalities is intuitively appealing for improving the performance of learning-based approaches. However, it is challenging to fully leverage different modalities due to practical challenges such as varying levels of noise and conflicts between modalities. Existing metho…
ViViT efficiently computes curvature for deep networks without approximations.
problem Efficiently computing curvature for deep networks without approximations.
method Leverages the GGN's low-rank structure without further approximations.
result ViViT allows for efficient computation of eigenvalues, eigenvectors, and directional derivatives.
DP-RandP improves privacy-utility tradeoff in DP-SGD by learning priors from random processes.
problem Improving the performance of differentially private stochastic gradient descent (DP-SGD) on private data.
method A three-phase approach that learns priors from images generated by random processes and transfers these priors to private data.
result New state-of-the-art accuracy on CIFAR10, CIFAR100, MedMNIST, and ImageNet for various privacy budgets.
Diffusion models outperform GANs in image synthesis quality.
problem Improving image synthesis quality using generative models.
method Developed and optimized diffusion models, and introduced classifier guidance for conditional synthesis.
result Achieved superior image synthesis quality with diffusion models, matching or exceeding GANs in FID scores.
We consider the problem of estimating the mean of a symmetric log-concave distribution under the constraint that only a single bit per sample from this distribution is available to the estimator. We study the mean squared error as a function of the sample size (and hence the number of bits). We consider three settings:…
When eliciting judgements from humans for an unknown quantity, one often has the choice of making direct-scoring (cardinal) or comparative (ordinal) measurements. In this paper we study the relative merits of either choice, providing empirical and theoretical guidelines for the selection of a measurement scheme. We pro…
When the cost of misclassifying a sample is high, it is useful to have an accurate estimate of uncertainty in the prediction for that sample. There are also multiple types of uncertainty which are best estimated in different ways, for example, uncertainty that is intrinsic to the training set may be well-handled by a B…
Logit regularization induces logit clustering, affecting classifier performance.
problem Understanding the mechanism of logit regularization in classification.
method Analysis of logit regularization in linear classification, proving logit clustering leads to Fisher's Linear Discriminant alignment.
result Logit regularization can halve critical sample complexity and induce robust generalization.
Deep learning takes advantage of large datasets and computationally efficient training algorithms to outperform other approaches at various machine learning tasks. However, imperfections in the training phase of deep neural networks make them vulnerable to adversarial samples: inputs crafted by adversaries with the int…
Paper proposes an unbiased optimization method for Bayesian experimental design.
problem Maximizing expected information gain in Bayesian experimental design.
method Randomized multilevel Monte Carlo (MLMC) method combined with stochastic gradient descent.
result An unbiased estimator for the gradient of expected information gain.
High throughput genetic sequencing arrays with thousands of measurements per sample and a great amount of related censored clinical data have increased demanding need for better measurement specific model selection. In this paper we establish strong oracle properties of nonconcave penalized methods for nonpolynomial (N…
Footfall based biometric system is perhaps the only person identification technique which does not hinder the natural movement of an individual. This is a clear edge over all other biometric systems which require a formidable amount of human intervention and encroach upon an individual's privacy to some extent or the o…
Recent progress in deep latent variable models has largely been driven by the development of flexible and scalable variational inference methods. Variational training of this type involves maximizing a lower bound on the log-likelihood, using samples from the variational posterior to compute the required gradients. Rec…
AMF-VI uses adaptive mixtures of flows for robust VI across diverse distributions.
problem Inconsistent behavior of single-flow models across different distributions.
method Sequential expert training of individual flows and adaptive global weight estimation via likelihood-driven updates.
result AMF-VI achieves lower negative log-likelihood and stable gains in transport metrics across various posterior families.
A new EM algorithm improves inference from large datasets.
problem Efficient inference in latent variable models with large datasets.
method Introduces SPIDER-EM, a novel EM algorithm using SPIDER estimator.
result Finite-time complexity bounds for smooth non-convex likelihood.
Spectral measurements reveal hidden representation geometry in language model training.
problem Hidden internal representation in language model training is hard to examine.
method Empirical protocol using activation covariance and per-sample gradient SVD spectra.
result Batch size affects representation geometry, and activation spectra predict token efficiency.
Energy-efficient DL inference for IoT devices reduces power consumption and improves performance.
problem Energy inefficiency in deep learning models for IoT devices.
method Energy-aware early exiting policy to balance energy consumption and inference accuracy.
result Accuracy and service rate improved up to 25% and 35% respectively.
SupSiam and SupBYOL improve supervised representation learning with ANCL.
problem Improving supervised representation learning with ANCL.
method Proposed supervised ANCL framework leveraging labels to avoid collapse.
result Supervised ANCL improves representation learning across various datasets and tasks.