Papers & research
Publication (listed in comprehensive documentation)
End-to-End Bangla AI for Solving Math Olympiad Problem Benchmark: Leveraging Large Language Model Using Integrated Approach
Add venue, year, authors, and a stable link (IEEE / ACM / arXiv / institution repository) when you publish the site.
Research directions (from your CV narrative)
AI/ML
- Retrieval quality and factual grounding in LLM systems
- Human-evaluation pipelines for applied RAG
- Educational AI and multilingual / low-resource settings
Cybersecurity
- Secure AI application architecture and attack-surface reduction
- Web app security and adversarial behavior in LLM-integrated systems
- Practical offensive/defensive learning via CTF
Future entries
| Type | Title / topic | Link |
|---|---|---|
| Preprint | Optional: expand CivilGPT or Curiosity evaluation into a paper | |
| Workshop | TBD |
Internal benchmarks (e.g. 16k QA Curiosity vs. ChatGPT) are strong engineering evidence but are not peer-reviewed publications unless you write them up formally.