Kancheti Sai Srinivas
 |
Kancheti Sai Srinivas |
Short Bio
I am a fifth-year Ph.D. student at Indian Institute of Technology Hyderabad. I have a dual-degree (Bachelors and Masters) in CSE from Indian Institute of Technology Madras. I am advised by Prof. Vineeth N Balasubramanian. You can find my CV here.
News
[2025] Our work titled “Semantic Alignment for Prompt-Tuning in Vision Language Models” has been accepted to TMLR openreview link
[2024] Awarded the Microsoft India Research PhD Award 2024
[2024] Our work titled “Fine-Grained Visual Recognition in the Age of Multimodal LLMs” was accepted at the NeurIPS 2024 Workshop: Adaptive Foundation Models
[2024] Presented our work “Interpretable Model Drift Detection” at CODS-COMAD 2024
[2023] Attended Google Research Week 2023
[2023] Attended ACML 2023 in Bangalore
[2021] Selected as a Prime Minister Research Fellow in the December 2020 cycle.
Research Interests
Large Multimodal Models
Algorithmic Recourse
Interpretable and Trustworthy ML
Causality in machine learning
Publications
Here is my Google Scholar
[2025] Semantic Alignment for Prompt-Tuning in Vision Language Models – K. Sai Srinivas*, Hari Chandana K*, Gowtham Reddy A, V. Balasubramanian, TMLR 2025 openreview link
[2024] Fine-Grained Visual Recognition in the Age of Multimodal LLMs – Hari Chandana K, K. Sai Srinivas, Gowtham Reddy A, V. Balasubramanian, NeurIPS 2024 Workshop: Adaptive Foundation Models link
[2024] Interpretable Model Drift Detection – Pranoy Panda, K. Sai Srinivas, V. Balasubramanian, Gaurav Sinha, CODS-COMAD ’24 link
[2023] A Simple Test-time Adaptation Method for Source-free Domain Generalization – Vaasudev Narayanan, K. Sai Srinivas, Sriranjani Ramakrishnan, V. Balasubramanian, ICLR 2023 Workshop link
[2022] Matching Learned Causal Effects of Neural Networks with Domain Priors – K. Sai Srinivas*, A. Gowtham Reddy*, V. Balasubramanian, Amit Sharma, ICML’22 link
Ongoing Projects
[2025] Finetuning VLMs under Complex Label Noise – Investigating parameter-efficient methods to adapt large VLMs to downstream tasks while handling noisy labels using the manifold hypothesis.
[2024] (under review) Human-in-the-Loop Algorithmic Recourse – Developing personalized recourses through iterative preference elicitation.
Education
PhD, Department of CSE, Indian Institute of Technology Hyderabad, Telangana, India, Jan. 2021 - Present
Dual-Degree (BTech & MTech), Department of CSE, Indian Institute of Technology Madras, Tamil Nadu, India, July. 2013 - July. 2019
Services
Reviewer: AAAI (2025,2024), AISTATS 2025, CVPR 2025, NeurIPS 2024, ECCV 2024.
Teaching Assistant: Foundations of Machine Learning (Fall 2022), Advanced Topics in ML (Spring 2024)