Martin Takáč

LoFT: Low-Rank Adaptation That Behaves Like Full Fine-Tuning

Newton Method Revisited: Global Convergence Rates up to $O(1/k^3)$ for Stepsize Schedules and Linesearch Procedures

Y-shaped Generative Flows

$\psi$DAG: Projected Stochastic Approximation Iteration for Linear DAG Structure Learning

Exponential Objective Decrease in Convex Setup is Possible! Gradient Descent Method Variants under $(L_0,L_1)$-Smoothness

Loss Transformation Invariance of the Damped Newton Methods

Preconditioned Norms: A Unified Framework for Steepest Descent, Quasi-Newton and Adaptive Methods

Through BabyAI Steps: Understanding and Evaluating Grounded Intelligence in LLMs

Selective Collaboration for Robust Federated Learning

Thinking like a CHEMIST: Combined Heterogeneous Embedding Model Integrating Structure and Tokens

Simple Stepsizes for Quasi-Newton Methods with Global Convergence Guarantees

Revisiting One-Shot Pruning with Scalable Second-Order Approximations

Expert or not? Assessing data quality in offline reinforcement learning

$\psi$DAG: Projected Stochastic Approximation Iteration for Linear DAG Structure Learning

From Risk to Uncertainty: Generating Predictive Uncertainty Measures via Bayesian Estimation

Methods with Local Steps and Random Reshuffling for Generally Smooth Non-Convex Federated Optimization

Uncovering the Spectral Bias in Diagonal State Space Models

OPTAMI: Global Superlinear Convergence of High-order Methods

Methods for Convex $(L_0,L_1)$-Smooth Optimization: Clipping, Acceleration, and Adaptivity

SANIA: Polyak-type Optimization Framework Leads to Scale Invariant Stochastic Algorithms

$MirrorCheck$ : Efficient Adversarial Defense for Vision-Language Models

Clipping Improves Adam and AdaGrad when the Noise Is Heavy-Tailed

FRUGAL: Memory-Efficient Optimization by Reducing State Overhead for Scalable Training

Exploring New Frontiers in Vertical Federated Learning: the Role of Saddle Point Reformulation

FRUGAL: Memory-Efficient Optimization by Reducing State Overhead for Scalable Training

Clipping Improves Adam-Norm and AdaGrad-Norm when the Noise Is Heavy-Tailed

Federated Learning Can Find Friends That Are Advantageous

$\psi$DAG : Projected Stochastic Approximation Iteration for DAG Structure Learning

Collaborative and Efficient Personalization with Mixtures of Adaptors

Broadening Discovery through Structural Models: Multimodal Combination of Local and Structural Properties for Predicting Chemical Features.

FedPeWS: Personalized Warmup via Subnetworks for Enhanced Heterogeneous Federated Learning

Adaptive Regularized Newton Methods with Inexact Hessian