Kiran Purohit (on job market)
@kiranpurohit08
PhD student in CSE at IIT Kharagpur. Trying to learn about deep learning faster than deep learning can learn about me.
ID: 777569982024523776
https://kiranpurohit.github.io/ 18-09-2016 18:08:25
67 Tweet
58 Followers
135 Following
Proud to see our EMNLP paper, EXPLORA: Efficient Exemplar Subset Selection for Complex Reasoning, being presented by my Prof. Sourangshu Bhattacharya, at the Ramakrishna Mission Vivekananda Educational and Research Institute (rkmvu.ac.in). Paper - aclanthology.org/2024.emnlp-maiโฆ
๐ฏ Honored to have attended the Google DeepMind Research Symposium 2025! A big shoutout to the Google DeepMind team for an unforgettable experience - from the stellar speaker lineup to the delicious food!๐ Abhilash Nandy Soumyadeep Roy Pretam Ray Payel Santra
Congratulations Kiran Purohit (on job market) for attending and presenting her work โEXPLORA: Efficient Exemplar Subset Selection for Complex Reasoningโ at Pingala Interactions in Computing (#PIC2025) held at Infosys Mysore campus! ๐
I had the incredible opportunity to attend ACM Pingala Interactions in Computing (PIC) 2025. This event brought together 50 young researchers from India and abroad, along with eminent laureates in Computing, fostering an inspiring research environment. Association for Computing Machinery #PIC2025
๐ ๐๐ฒ๐ป๐๐๐ฟ๐ ๐ถ๐ ๐ป๐ผ๐ ๐๐ป๐ผ๐๐ด๐ต . . . ๐๐ข๐ฑ๐ต๐ช๐ฐ๐ฏ ๐๐ณ๐ฆ๐ฅ๐ช๐ต: Kishalay Das (KD) ๐๐ฏ๐ด๐ฑ๐ช๐ณ๐ฆ๐ฅ ๐ฃ๐บ ๐๐ฐ๐ถ๐ณ๐ข๐ท ๐๐ข๐ฏ๐จ๐ถ๐ญ๐บ'๐ด ๐ข๐ถ๐ต๐ฐ๐ฃ๐ช๐ฐ๐จ๐ณ๐ข๐ฑ๐ฉ๐บ #Research #AI #MachineLearning #PhDLife
๐Iโm pleased to share that I have joined ๐๐ฑ๐ผ๐ฏ๐ฒ ๐ฅ๐ฒ๐๐ฒ๐ฎ๐ฟ๐ฐ๐ต as a Research PhD Intern ๐ฏThe work will focus on ๐น๐ฎ๐ป๐ด๐๐ฎ๐ด๐ฒ ๐บ๐ผ๐ฑ๐ฒ๐น ๐ฎ๐น๐ถ๐ด๐ป๐บ๐ฒ๐ป๐. This aligns well with my ongoing research interests in making large language models reliable and human-centered.
๐ Excited to share the preprint and code for our ICML 2025 paper. ๐ Preprint: "Sample Efficient Demonstration Selection for ICL" ๐ arxiv.org/pdf/2506.08607 ๐ป Code: ๐ github.com/kiranpurohit/Cโฆ ๐ Project Page: ๐ kiranpurohit.github.io/CASE-Bandit/ venktesh, Sourangshu Bhattacharya, Avishek Anand
Can we compose skills for solving a task by principled selection of exemplars to use in In-Context Learning. We answer this question in our ICML 2025 paper with Kiran Purohit (on job market) Avishek Anand Sourangshu Bhattacharya ๐ Paper: arxiv.org/abs/2506.08607 ๐งโ๐ป Code: github.com/kiranpurohit/Cโฆ (1/5)
If you are attending ICML Conference, catch our poster tomorrow (16th July) at the Poster session (4:30 p.m. PDT โ 7 p.m. PDT)! ๐ฅ Video: recorder-v3.slideslive.com/#/share?share=โฆ ๐ Paper: openreview.net/pdf?id=cuqvlLBโฆ ๐ป Code: github.com/kiranpurohit/Cโฆ venktesh Sourangshu Bhattacharya Avishek Anand
๐ถ๐๐ ๐๐๐๐ ๐๐๐๐๐๐ ๐๐ ๐ซ๐. ๐ฒ๐๐๐๐ ๐ท๐๐๐๐๐๐ โ ๐ท๐๐ซ ๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐ ! ๐๐ ๐ฎ๐๐๐๐๐๐๐ ๐๐ ๐ท๐๐๐. Sourangshu Bhattacharya ๐๐๐ ๐๐๐ ๐๐๐๐ ๐๐๐๐ ๐๐๐๐๐๐๐๐๐๐ ๐๐๐๐ ๐๐๐๐๐๐๐.
We bid farewell to Kiran Purohit (on job market) an invaluable member of CNeRG since 2020. Her dedication has left a lasting impact. Wishing you great success in your new role as a senior researcher-2 at Fujitsu Research. All the very best in all your future!๐
Itโs hard to say goodbye to a place thatโs been a home away from home. โค๏ธ IIT Kharagpur๐ฎ๐ณ and CNeRG IIT KGP shaped me, challenged me, and gifted me memories, friendships & learnings Iโll always cherish. Will miss you all. Cheers to new beginnings! ๐ฅ