Picture for Sergey Troshin

Sergey Troshin

HSE University, Russia

ARM: Efficient Guided Decoding with Autoregressive Reward Models

Add code
Jul 05, 2024
Viaarxiv icon

CodeBPE: Investigating Subtokenization Options for Large Language Model Pretraining on Source Code

Add code
Aug 01, 2023
Viaarxiv icon

SantaCoder: don't reach for the stars!

Add code
Jan 09, 2023
Viaarxiv icon

Probing Pretrained Models of Source Code

Add code
Feb 16, 2022
Figure 1 for Probing Pretrained Models of Source Code
Figure 2 for Probing Pretrained Models of Source Code
Figure 3 for Probing Pretrained Models of Source Code
Figure 4 for Probing Pretrained Models of Source Code
Viaarxiv icon

Machine Learning Methods for Spectral Efficiency Prediction in Massive MIMO Systems

Add code
Dec 29, 2021
Figure 1 for Machine Learning Methods for Spectral Efficiency Prediction in Massive MIMO Systems
Figure 2 for Machine Learning Methods for Spectral Efficiency Prediction in Massive MIMO Systems
Figure 3 for Machine Learning Methods for Spectral Efficiency Prediction in Massive MIMO Systems
Figure 4 for Machine Learning Methods for Spectral Efficiency Prediction in Massive MIMO Systems
Viaarxiv icon

A Simple Approach for Handling Out-of-Vocabulary Identifiers in Deep Learning for Source Code

Add code
Oct 23, 2020
Figure 1 for A Simple Approach for Handling Out-of-Vocabulary Identifiers in Deep Learning for Source Code
Figure 2 for A Simple Approach for Handling Out-of-Vocabulary Identifiers in Deep Learning for Source Code
Viaarxiv icon

Empirical Study of Transformers for Source Code

Add code
Oct 15, 2020
Figure 1 for Empirical Study of Transformers for Source Code
Figure 2 for Empirical Study of Transformers for Source Code
Figure 3 for Empirical Study of Transformers for Source Code
Figure 4 for Empirical Study of Transformers for Source Code
Viaarxiv icon