Indian Institute of Technology, Hyderabad
Jan 2023 - Dec 2025
Master of Technology in Computer Science & Engineering (By Research) - CGPA: (9.76/10)
Rahul Vigneswaran's Resume
Rahul Vigneswaran
Masters in Computer Science & Engineering (By Research), Reliance Foundation Fellow
About
AI Researcher at IIT Hyderabad focusing on Long-Tailed Class Incremental Learning, Human-in-the-Loop AI, and LLM applications. Published in TMLR, AAAI, and ICVGIP conferences with papers under review at AAAI, ICLR, and EMNLP.
Education
Publications
Class Incremental Learning Free of Unrealistic Assumptions
Rahul Vigneswaran, Divya, Digvijay, Chandana, Vineeth N Balasubramanian
AAAI 2026
Under Review
Tackling Long-Tailed Class Incremental Learning
Rahul Vigneswaran, Chandana, Vineeth N Balasubramanian
ICLR 2026
Under Review
Structure Thinking for enhanced reasoning in LLMs
Anubhav, Rahul Vigneswaran, Stanley, Anish, Promod
EMNLP 2025
Under Review
A Deeper Look at the Hessian Eigen Spectrum of Deep Neural Networks and its Applications to Regularization
Adepu Ravi Shankar*, Yash Khasbage*, Rahul Vigneswaran, Vineeth N Balasubramanian
AAAI 2021
Published
Research Experience
Applied Scientist Intern
Mar 2025 - Aug 2025
Amazon, India
Managers: Promod Yenigalla, Anish Nediyanchath, Anubhav Shrimal
- Developed a new capability for Amazon's internal Agentic assistant (SAPIEN), enabling insight exploration across databases and user files through automatic table identification (≈89% accuracy), context-aware clarification handling, and workflow initiation; demoed to leadership
- Reduced manual testing effort of SAPIEN by building tools like Golden Dataset Generator and Gamma Testing Framework, ensuring scalable reliability across use cases
- Secured a Top-3 finish (only intern to do so) at Amazon's internal Hackathon with Promptinator-3000, an automated prompt and dataset generation framework, earning recognition from leadership
- Work under review at AMLC'25 (Amazon Internal) and EMNLP'25
Research Assistant
Jan 2023 - Present
IIT Hyderabad, India
Advisor: Dr. Vineeth NB
- Proposed a new realistic setup in Continual Learning that is free of assumption and developed a novel method using adapters to tackle it
- Developed a novel method to tackle Transitioning Head problem in Long-Tailed Class Incremental Learning via Early Knowledge Transfer, achieving state-of-the-art results
- Created a human-in-the-loop recourse framework that integrates user feedback, generating personalized counterfactuals and enhancing user satisfaction and transparency
- Two works under review at AAAI'25 and ICLR'26. One work published at TMLR'24/25
Research Intern
July 2019 - Jan 2023
IIT Hyderabad, India
Advisors: Dr. Vineeth NB (IIT-H) & Dr. Makarand Tapaswi (IIIT-H)
- Developed TailCalibX, a feature generation technique for Long-Tailed classification that uses calibrated distributions to boost performance on imbalanced datasets, setting a new state-of-the-art
- Created a Hessian-based regularization method that improves generalization by leveraging the similarity between layerwise and overall Hessians, enhancing regularizer efficiency
- Published in AAAI'21 and ICVGIP'21
Research projects
Integration Testing for Stochastic AI systems
Developed an LLM-based, persona-driven multi-turn evaluation framework for stochastic AI systems, simulating real user behaviors and enabled headless CI/CD integration to produce automated pass/fail signals.
- LLMs
- Testing
- Persona-driven evaluation
Adversarially Coupled Prompt & Dataset Generator
Developed an adversarial co-evolution framework coupling a prompt generator and dataset generator with access to intermediate reasoning steps, achieving a 21× efficiency gain.
- LLMs
- Prompt Engineering
- Adversarial Co-evolution
AWARE: Adaptive Wear-levelling and Attack Re-mapping Engine
A framework that enhances NVM durability and security in LLCs by combining adaptive wear-leveling and attack mitigation through intelligent remapping.
- Computer Architecture
- NVM
- Security
LLM Hardware Optimizations
Analyzed hardware and software optimizations to improve Large Language Models, identifying gaps like fragmented benchmarking and advocating for unified solutions.
- Computer Architecture
- LLMs
- Hardware Acceleration
Neural Collapse in Long-Tailed Continual Learning
Uncovered and addressed key limitations in existing theoretical frameworks for analyzing Neural Collapse in continual learning, extending their applicability to Long-Tailed Continual Learning.
- Deep Learning
- Theoretical ML
- Long-Tailed Learning
TARM: Token Averaging Recurrent Memory Transformers
A novel method using exponential moving average on memory tokens to boost memory capacity in Recurrent Memory Transformers, enhancing long-term dependency capture and training stability.
- Transformers
- Memory Networks
- Deep Learning
Achievements
Reliance Foundation Postgraduate Scholarship
Awarded to Top 100 Students Nationwide. Scholarship Value: 6 Lakhs
2023-2025
Amazon's Internal Hackathon
Top-3. Only Intern Finalist, competing against well-seasoned Applied Scientists & SDEs
2025
TiDeL Hackathon
First Place
2024
Amazon Machine Learning Summer School
Selected Participant
2024
Volunteering
Research
- Reviewer: AAAI 2026, ECCV 2022
- Sub-Reviewer: CVPR 2023, ICLR 2021, IJCAI 2020, WACV 2023, SDM 2021
- Student Volunteer: ACML 2022, ICML 2020
Academic TAships
- Deep Learning for Computer Vision (NPTEL) (2024, 2020)
- AI and Emerging Technologies (TalentSprint + IIT Hyderabad) (2024, 2023, 2022)
- Effective Teaching of Machine Learning (CSEDU IIT Delhi) (2022, 2021)
- Reinforcement Learning (AI 3000 / CS 5500) (IIT Hyderabad) (2022)
- Advanced Topics in Machine Learning (AI 2100 / CS 6360) (IIT Hyderabad) (2021)
References
Dr. Vineeth N. Balasubramanian
Principal Researcher, Microsoft Research, India | Professor, IIT-H, India
Promod Yenigalla
Sr. Applied Science Manager, Amazon, USA
Anish Nediyanchath
Sr. Applied Scientist, Amazon, India
Anubhav Shrimal
Sr. Applied Scientist, Amazon, USA