# research.md — Academic Research & Publications

This document details the academic papers, book chapters, and research investigations authored by Suyash Shukla.

---

## 1. Published Research Paper

### "Bias Detection and Mitigation in Large Language Models: A Fairness-Driven Approach"
- **Publisher:** Taylor & Francis (CRC Press)
- **Year of Publication:** 2025
- **Authors:** Suyash Shukla, et al.
- **Subject Areas:** Natural Language Processing, Algorithmic Fairness, Ethical AI, Large Language Models
- **Abstract & Focus:**
  Contemporary autoregressive large language models exhibit non-trivial demographic, cultural, and representational disparities inherited from uncurated pre-training corpora. This research formalizes a multi-dimensional evaluation benchmark evaluating stereotyping and disparity across protected demographic groups. It introduces a post-hoc counterfactual mitigation strategy that adjusts generation logits at inference time without requiring expensive full-model retraining, preserving downstream task accuracy while reducing disparate toxicity and stereotyping scores.

---

## 2. Book Chapters & Forthcoming Scholarship

### Applied Data Science in Practical STEM Pedagogy
- **Co-authored:** 3 Book Chapters & 2 Research Papers (in process and under peer review)
- **Focus:**
  - Architecting high-throughput browser-based laboratory simulations using WebSockets and containerized FastAPI worker pods.
  - Designing automated grading algorithms with code and calculation verification.
  - Real-time measurement distribution across multi-tenant academic environments.

---

## 3. Core Research Interests
1. **Model Safety & Reliability:** Deterministic verification of agentic tool-calling outputs using state-graph invariants.
2. **Algorithmic Fairness:** Counterfactual testing and debiasing frameworks for multilingual text generation.
3. **Efficient Model Alignment:** Parameter-efficient fine-tuning (PEFT, LoRA/QLoRA) and reward model calibration for open-weights foundation models.

---

## 4. Related Resources
- [Knowledge Home](/knowledge/me.md)
- [Education](/knowledge/education.md)
- [Projects](/knowledge/projects.md)
- [HTML Research Page](/research)