Tatiana Likhomanenko

collaborators

9 papers

cs.CL2026

Scaling Properties of Continuous Diffusion Spoken Language Models

Jason Ramapuram, Eeshan Gunesh Dhekane, Amitis Shidani +6

Speech-only spoken language models (SLMs) lag behind text and text-speech models in performance, with recent discrete autoregressive (AR) SLMs indicating significant computational…

cs.LG2025

Enabling Differentially Private Federated Learning for Speech Recognition: Benchmarks, Adaptive Optimizers and Gradient Clipping

Martin Pelikan, Sheikh Shams Azam, Vitaly Feldman +4

While federated learning (FL) and differential privacy (DP) have been extensively studied, their application to automatic speech recognition (ASR) remains largely unexplored due to…

cs.LG2025

DiceHuBERT: Distilling HuBERT with a Self-Supervised Learning Objective

Hyung Gun Chi, Zakaria Aldeneh, Tatiana Likhomanenko +5

We introduce DiceHuBERT, a knowledge distillation framework for compressing HuBERT, a widely used self-supervised learning (SSL)-based speech foundation model. Unlike existing dist…

hep-ex2025

Evidence for CP violation in $B^{+} \to p p ˉ K^{+}$ decays

LHCb collaboration, R. Aaij, B. Adeva +698

Three-body $B^{+} \to p p K^{+}$ and $B^{+} \to p p I ¨ ^{+}$ decays are studied using a data sample corresponding to an integrated luminosity of 3.0 $f b^{- 1}$ collected by…

hep-ex2025

Measurements of charm mixing and $C P$ violation using $D^{0} \to K^{\pm} I ¨ ^{\mp}$ decays

LHCb collaboration, R. Aaij, B. Adeva +766

Measurements of charm mixing and $C P$ violation parameters from the decay-time-dependent ratio of $D^{0} \to K^{+} I ¨ ^{-}$ to $D^{0} \to K^{-} I ¨ ^{+}$ decay rates and the charge-conjugat…

cs.LG2025

Theory, Analysis, and Best Practices for Sigmoid Self-Attention

Jason Ramapuram, Federico Danieli, Eeshan Dhekane +8

Attention is a key part of the transformer architecture. It is a sequence-to-sequence mapping that transforms each sequence element into a weighted sum of values. The weights are t…

Scaling Properties of Continuous Diffusion Spoken Language Models

Enabling Differentially Private Federated Learning for Speech Recognition: Benchmarks, Adaptive Optimizers and Gradient Clipping

DiceHuBERT: Distilling HuBERT with a Self-Supervised Learning Objective

Evidence for CP violation in B+→ppˉ​K+ decays

Measurements of charm mixing and CP violation using D0→K±I¨∓ decays

Theory, Analysis, and Best Practices for Sigmoid Self-Attention

Evidence for CP violation in $B^{+} \to p p ˉ K^{+}$ decays

Measurements of charm mixing and $C P$ violation using $D^{0} \to K^{\pm} I ¨ ^{\mp}$ decays