Research
// Intellectual curiosity and current explorations
Neural Architecture Search
Automating the design of efficient neural networks for resource-constrained environments. Exploring differentiable architecture search and zero-shot proxy tasks.
Key Papers
DARTS: Differentiable Architecture Search
Once-for-All: Train One Network and Specialize it for Efficient Deployment
Retrieval-Augmented Generation
Building intelligent systems that combine parametric knowledge with external retrieval. Focused on chunking strategies, embedding optimization, and query rewriting.
Key Papers
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
Dense Passage Retrieval for Open-Domain Question Answering
Agentic AI Systems
Designing autonomous agents that can plan, reason, and execute complex tasks. Exploring LangGraph for multi-agent orchestration and tool use.
Key Papers
ReAct: Synergizing Reasoning and Acting in Language Models
LangGraph: Building Stateful, Multi-Agent Applications
Computer Vision
Efficient vision transformers and self-supervised learning for visual representation. Interested in medical imaging and autonomous systems applications.
Key Papers
An Image is Worth 16x16 Words: Transformers for Image Recognition
DINO: Emerging Properties in Self-Supervised Vision Transformers
Future Directions
- →Exploring multimodal AI systems that combine vision, language, and structured reasoning
- →Building self-improving agents that learn from their own execution traces
- →Investigating efficient fine-tuning methods for domain-specific LLMs