Alert button
Picture for Arturs Backurs

Arturs Backurs

Alert button

Privately Aligning Language Models with Reinforcement Learning

Oct 25, 2023
Fan Wu, Huseyin A. Inan, Arturs Backurs, Varun Chandrasekaran, Janardhan Kulkarni, Robert Sim

Viaarxiv icon

Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping

Dec 03, 2022
Jiyan He, Xuechen Li, Da Yu, Huishuai Zhang, Janardhan Kulkarni, Yin Tat Lee, Arturs Backurs, Nenghai Yu, Jiang Bian

Figure 1 for Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping
Figure 2 for Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping
Figure 3 for Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping
Figure 4 for Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping
Viaarxiv icon

Unveiling Transformers with LEGO: a synthetic reasoning task

Jun 09, 2022
Yi Zhang, Arturs Backurs, Sébastien Bubeck, Ronen Eldan, Suriya Gunasekar, Tal Wagner

Figure 1 for Unveiling Transformers with LEGO: a synthetic reasoning task
Figure 2 for Unveiling Transformers with LEGO: a synthetic reasoning task
Figure 3 for Unveiling Transformers with LEGO: a synthetic reasoning task
Figure 4 for Unveiling Transformers with LEGO: a synthetic reasoning task
Viaarxiv icon

Differentially Private Model Compression

Jun 03, 2022
Fatemehsadat Mireshghallah, Arturs Backurs, Huseyin A Inan, Lukas Wutschitz, Janardhan Kulkarni

Figure 1 for Differentially Private Model Compression
Figure 2 for Differentially Private Model Compression
Figure 3 for Differentially Private Model Compression
Figure 4 for Differentially Private Model Compression
Viaarxiv icon

Differentially Private Fine-tuning of Language Models

Oct 13, 2021
Da Yu, Saurabh Naik, Arturs Backurs, Sivakanth Gopi, Huseyin A. Inan, Gautam Kamath, Janardhan Kulkarni, Yin Tat Lee, Andre Manoel, Lukas Wutschitz, Sergey Yekhanin, Huishuai Zhang

Figure 1 for Differentially Private Fine-tuning of Language Models
Figure 2 for Differentially Private Fine-tuning of Language Models
Figure 3 for Differentially Private Fine-tuning of Language Models
Figure 4 for Differentially Private Fine-tuning of Language Models
Viaarxiv icon

Faster Kernel Matrix Algebra via Density Estimation

Feb 16, 2021
Arturs Backurs, Piotr Indyk, Cameron Musco, Tal Wagner

Figure 1 for Faster Kernel Matrix Algebra via Density Estimation
Figure 2 for Faster Kernel Matrix Algebra via Density Estimation
Viaarxiv icon

Generating (Formulaic) Text by Splicing Together Nearest Neighbors

Jan 29, 2021
Sam Wiseman, Arturs Backurs, Karl Stratos

Figure 1 for Generating (Formulaic) Text by Splicing Together Nearest Neighbors
Figure 2 for Generating (Formulaic) Text by Splicing Together Nearest Neighbors
Figure 3 for Generating (Formulaic) Text by Splicing Together Nearest Neighbors
Figure 4 for Generating (Formulaic) Text by Splicing Together Nearest Neighbors
Viaarxiv icon

Impossibility Results for Grammar-Compressed Linear Algebra

Oct 27, 2020
Amir Abboud, Arturs Backurs, Karl Bringmann, Marvin Künnemann

Figure 1 for Impossibility Results for Grammar-Compressed Linear Algebra
Figure 2 for Impossibility Results for Grammar-Compressed Linear Algebra
Viaarxiv icon

Active Local Learning

Sep 04, 2020
Arturs Backurs, Avrim Blum, Neha Gupta

Viaarxiv icon