Alert button
Picture for Janardhan Kulkarni

Janardhan Kulkarni

Alert button

Differentially Private Training of Mixture of Experts Models

Feb 11, 2024
Pierre Tholoniat, Huseyin A. Inan, Janardhan Kulkarni, Robert Sim

Viaarxiv icon

TinyGSM: achieving >80% on GSM8k with small language models

Dec 14, 2023
Bingbin Liu, Sebastien Bubeck, Ronen Eldan, Janardhan Kulkarni, Yuanzhi Li, Anh Nguyen, Rachel Ward, Yi Zhang

Viaarxiv icon

Classification with Partially Private Features

Dec 11, 2023
Zeyu Shen, Anilesh Krishnaswamy, Janardhan Kulkarni, Kamesh Munagala

Viaarxiv icon

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

Assessing Privacy Risks in Language Models: A Case Study on Summarization Tasks

Oct 20, 2023
Ruixiang Tang, Gord Lueck, Rodolfo Quispe, Huseyin A Inan, Janardhan Kulkarni, Xia Hu

Viaarxiv icon

Privacy-Preserving In-Context Learning with Differentially Private Few-Shot Generation

Sep 21, 2023
Xinyu Tang, Richard Shin, Huseyin A. Inan, Andre Manoel, Fatemehsadat Mireshghallah, Zinan Lin, Sivakanth Gopi, Janardhan Kulkarni, Robert Sim

Figure 1 for Privacy-Preserving In-Context Learning with Differentially Private Few-Shot Generation
Figure 2 for Privacy-Preserving In-Context Learning with Differentially Private Few-Shot Generation
Figure 3 for Privacy-Preserving In-Context Learning with Differentially Private Few-Shot Generation
Figure 4 for Privacy-Preserving In-Context Learning with Differentially Private Few-Shot Generation
Viaarxiv icon

Differentially Private Synthetic Data via Foundation Model APIs 1: Images

May 24, 2023
Zinan Lin, Sivakanth Gopi, Janardhan Kulkarni, Harsha Nori, Sergey Yekhanin

Figure 1 for Differentially Private Synthetic Data via Foundation Model APIs 1: Images
Figure 2 for Differentially Private Synthetic Data via Foundation Model APIs 1: Images
Figure 3 for Differentially Private Synthetic Data via Foundation Model APIs 1: Images
Figure 4 for Differentially Private Synthetic Data via Foundation Model APIs 1: Images
Viaarxiv icon

Selective Pre-training for Private Fine-tuning

May 23, 2023
Da Yu, Sivakanth Gopi, Janardhan Kulkarni, Zinan Lin, Saurabh Naik, Tomasz Lukasz Religa, Jian Yin, Huishuai Zhang

Figure 1 for Selective Pre-training for Private Fine-tuning
Figure 2 for Selective Pre-training for Private Fine-tuning
Figure 3 for Selective Pre-training for Private Fine-tuning
Figure 4 for Selective Pre-training for Private Fine-tuning
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

When Does Differentially Private Learning Not Suffer in High Dimensions?

Jul 09, 2022
Xuechen Li, Daogao Liu, Tatsunori Hashimoto, Huseyin A. Inan, Janardhan Kulkarni, Yin Tat Lee, Abhradeep Guha Thakurta

Figure 1 for When Does Differentially Private Learning Not Suffer in High Dimensions?
Figure 2 for When Does Differentially Private Learning Not Suffer in High Dimensions?
Figure 3 for When Does Differentially Private Learning Not Suffer in High Dimensions?
Figure 4 for When Does Differentially Private Learning Not Suffer in High Dimensions?
Viaarxiv icon