Machine Learning

token budget is becoming part of my agent workflow design

I think token budget is becoming part of agent workflow design. If every run feels expensive, people under-test. They save quota, overthi...

Improving AI models’ ability to explain their predictions

New technique makes AI models leaner and faster while they’re still learning

[2604.04342] Generative models for decision-making under distributional shift

Abstract page for arXiv paper 2604.04342: Generative models for decision-making under distributional shift

[2604.04316] How Long short-term memory artificial neural network, synthetic data, and fine-tuning improve the classification of raw EEG data

Abstract page for arXiv paper 2604.04316: How Long short-term memory artificial neural network, synthetic data, and fine-tuning improve t...

[2604.04313] Convolutional Neural Network and Adversarial Autoencoder in EEG images classification

Abstract page for arXiv paper 2604.04313: Convolutional Neural Network and Adversarial Autoencoder in EEG images classification

[2604.04290] DAGAF: A directed acyclic generative adversarial framework for joint structure learning and tabular data synthesis

Abstract page for arXiv paper 2604.04290: DAGAF: A directed acyclic generative adversarial framework for joint structure learning and tab...

[2604.04287] Entropy, Disagreement, and the Limits of Foundation Models in Genomics

Abstract page for arXiv paper 2604.04287: Entropy, Disagreement, and the Limits of Foundation Models in Genomics

[2604.04261] APPA: Adaptive Preference Pluralistic Alignment for Fair Federated RLHF of LLMs

Abstract page for arXiv paper 2604.04261: APPA: Adaptive Preference Pluralistic Alignment for Fair Federated RLHF of LLMs

[2604.04255] Towards Unveiling Vulnerabilities of Large Reasoning Models in Machine Unlearning

Abstract page for arXiv paper 2604.04255: Towards Unveiling Vulnerabilities of Large Reasoning Models in Machine Unlearning

[2604.04241] Learning An Interpretable Risk Scoring System for Maximizing Decision Net Benefit

Abstract page for arXiv paper 2604.04241: Learning An Interpretable Risk Scoring System for Maximizing Decision Net Benefit

[2604.04239] Good Rankings, Wrong Probabilities: A Calibration Audit of Multimodal Cancer Survival Models

Abstract page for arXiv paper 2604.04239: Good Rankings, Wrong Probabilities: A Calibration Audit of Multimodal Cancer Survival Models

[2604.04231] Subspace Control: Turning Constrained Model Steering into Controllable Spectral Optimization

Abstract page for arXiv paper 2604.04231: Subspace Control: Turning Constrained Model Steering into Controllable Spectral Optimization

[2604.04230] Three Phases of Expert Routing: How Load Balance Evolves During Mixture-of-Experts Training

Abstract page for arXiv paper 2604.04230: Three Phases of Expert Routing: How Load Balance Evolves During Mixture-of-Experts Training

[2604.04199] Which Leakage Types Matter?

Abstract page for arXiv paper 2604.04199: Which Leakage Types Matter?

[2604.04195] Stable and Privacy-Preserving Synthetic Educational Data with Empirical Marginals: A Copula-Based Approach

Abstract page for arXiv paper 2604.04195: Stable and Privacy-Preserving Synthetic Educational Data with Empirical Marginals: A Copula-Bas...

[2604.04175] Uncertainty-Aware Foundation Models for Clinical Data

Abstract page for arXiv paper 2604.04175: Uncertainty-Aware Foundation Models for Clinical Data

[2604.04155] The Geometric Alignment Tax: Tokenization vs. Continuous Geometry in Scientific Foundation Models

Abstract page for arXiv paper 2604.04155: The Geometric Alignment Tax: Tokenization vs. Continuous Geometry in Scientific Foundation Models

[2604.04107] Physical Sensitivity Kernels Can Emerge in Data-Driven Forward Models: Evidence From Surface-Wave Dispersion

Abstract page for arXiv paper 2604.04107: Physical Sensitivity Kernels Can Emerge in Data-Driven Forward Models: Evidence From Surface-Wa...

[2604.04101] Restless Bandits with Individual Penalty Constraints: A New Near-Optimal Index Policy and How to Learn It

Abstract page for arXiv paper 2604.04101: Restless Bandits with Individual Penalty Constraints: A New Near-Optimal Index Policy and How t...

[2604.04090] Fine-grained Analysis of Stability and Generalization for Stochastic Bilevel Optimization

Abstract page for arXiv paper 2604.04090: Fine-grained Analysis of Stability and Generalization for Stochastic Bilevel Optimization

[2604.04087] ArrowFlow: Hierarchical Machine Learning in the Space of Permutations

Abstract page for arXiv paper 2604.04087: ArrowFlow: Hierarchical Machine Learning in the Space of Permutations

[2604.04037] Geometric Limits of Knowledge Distillation: A Minimum-Width Theorem via Superposition Theory

Abstract page for arXiv paper 2604.04037: Geometric Limits of Knowledge Distillation: A Minimum-Width Theorem via Superposition Theory