Himanshu Sharma | GitHub Portfolio

Projects

Multi-Agent Multi-Arm Bandits with Minimum Reward Guarantee Fairness

Developed REWARDFAIRUCB, a UCB-based algorithm tailored for the multi-agent variant of the stochastic multi-armed bandit problem.

View on GitHub

Rumour Verification Based on Evidence from Authorities

Designed a joint training framework for Retrieval-Augmented Classification (RAG-based) leveraging authoritative Twitter evidence.

View on GitHub

Automated Cold Email Generation System for Service Companies

Constructed a complete ML pipeline integrating Llama 3.1 LLM for automated email generation.

View on GitHub

Publications

Multi-agent Multi-armed Bandits with Minimum Reward Guarantee Fairness - Accepted at AAMAS 2025
Retrieval Augmented Classification using Differentiable Top-K Operator for Rumor Verification - CLEF 2024

About Me

I am a researcher in Artificial Intelligence & Machine Learning, currently pursuing M.Tech by Research at IIT Hyderabad. I have experience in multi-agent systems, bandit algorithms, and NLP projects.

Projects

Multi-Agent Multi-Arm Bandits with Minimum Reward Guarantee Fairness

Rumour Verification Based on Evidence from Authorities

Automated Cold Email Generation System for Service Companies

Publications

About Me

Contact