Iryna Gurevych

Revela: Dense Retriever Learning via Language Modeling

Variational Model Merging for Pareto Front Estimation in Multitask Finetuning

ObscuraCoder: Powering Efficient Code LM Pre-Training Via Obfuscation Grounding

Differentially Private Steering for Large Language Model Alignment

How to Weight Multitask Finetuning? Fast Previews via Model Merging

Uncertainty-Aware Decoding with Minimum Bayes Risk

Early Period of Training Impacts Adaptation for Out-of-Distribution Generalization: An Empirical Study

Fine-Tuning with Divergent Chains of Thought Boosts Reasoning Through Self-Correction in Language Models

An Efficient Quantum Classifier Based on Hamiltonian Representations