Ali
Aghababaei

Explainable NLP with Integrated Gradients

Fine-tuned BERT, RoBERTa, and BERTAgent on the Slurs dataset, women's accounts of verbal harassment labeled for shame, fear, and sexual objectification. Applied Integrated Gradients via Captum for word-level attribution, benchmarked against Sequential IG, DeepLIFT, and GradientSHAP using comprehensiveness, sufficiency, and approximation error. Developed a custom rendering pipeline with SpaCy to handle negations and subword tokens. Validated findings with social psychologists, and applied the same method to a small-dataset scenario using deliberate overfitting to surface class-defining keywords. arXiv preprint (2025), submitted for journal publication.

Research Paper Q&A Agent

An agentic AI system built with LangGraph that answers research questions by searching arXiv live, retrieving relevant paper chunks, detecting contradictions between sources, and generating grounded answers with inline citations. The agent reasons and loops: if initial search results are insufficient, it reformulates queries and retries up to 3 times before gracefully exiting. Evaluated with RAGAS across 20 queries: Faithfulness 0.921, Answer Relevancy 0.885.

Key engineering decisions: per-paper ChromaDB retrieval with global re-ranking to prevent volume bias, session-isolated vector collections via UUID namespacing, dynamic query contextualization to bridge the conversational-to-academic embedding gap, and a gated RAGAS evaluator that runs offline to keep response time acceptable in the live demo.

Dutch Constitution Q&A (RAG Pipeline)

End-to-end Retrieval-Augmented Generation system that answers questions about the Dutch Constitution (2023) with Article citations, grounded entirely in the source document. Includes RAGAS evaluation, MLflow experiment tracking, and a documented engineering iteration: identified chunk boundary bleeding through tracked runs, replaced character-based splitting with structure-aware Article-level chunking, and improved Faithfulness from 0.55 to 0.86 (+0.31). Pipeline: PDF ingestion → article-aware split → MiniLM embeddings → ChromaDB → Llama 3.1 via Groq. Served via FastAPI REST API and Gradio UI on HuggingFace Spaces. GitHub Actions CI/CD pipeline runs 13 pytest tests on every push and auto-deploys on success.