Projects
Research artifacts and controlled experiments in retrieval, grounded generation, agent evaluation, failure analysis, and reproducible machine learning.
Featured3 projects
msmarco-genqa
activeReproducible retrieve → rerank → generate pipeline on MS MARCO with paired-bootstrap evaluation and a manifest-enforced reproducibility contract.
View project
rag-observatory
activeResearch prototype for trace-based RAG observability and failure analysis.
View project
Q-Learning Exploitability
maintainedExact agent evaluation and failure analysis for self-play Q-learning.
View project
Related research1 project